The Speed of the Librem 5 | Hacker News Reader

The Speed of the Librem 5

puri.sm

> In 2017, Apple admitted to intentionally slowing down older iPhones. The claim was this would increase your phone’s lifespan. But really, it made your once snappy phone feel much more sluggish, leading some to upgrade their phones to a newer model with marginally better specs

I doubt any of their marketing people are reading this but outright lying like that is not how you build trust in your products. They’re referring to the phone throttling its power draw when a degraded battery could no longer provide full output, and the alternative would be shutting down suddenly at a time when the phone was doing something demanding like playing a game. Other devices do this as well for the same reason, and anyone selling battery powered devices has no excuse for not knowing this. A company which shipped a phone in 2020 with 5 hours of battery life when new must be painfully aware of the challenges.

Similarly, anyone buying another phone rather than a much cheaper battery replacement is probably doing so because they want the additional performance, camera quality, etc. or simply normal wear and tear has added up over time and the repair costs are closer to replacement.

A company delivering a phone with performance slower than competing devices made a decade earlier similarly has no business claiming that the gains for most users are marginal just because their hardware has lagged far behind. There’s a great example of deceptive marketing here:

> The Librem 5 utilizes a quad-core arm based CPU, where Purism is the first company to include this CPU in a mobile phone

https://puri.sm/pages/iphone-vs-librem-5/

Someone who doesn’t know that quad core ARM CPUs started shipping in phones around 2011 would not understand that to mean “nobody else uses this specific CPU for phones” rather than what they’re hoping the reader will think: “this thing isn’t slower than an iPhone 4S” The security claims are similarly broad and designed to be read as more persuasive than is actually warranted.

This is disappointing because I like the idea of more open hardware, and it’s certainly the case that a lot of phones are thrown out because they stop getting support, but the more I read their copy the less I trust this company. They’re trying to make an argument for why you shouldn’t hold slower hardware against them, but they way they do it makes me less inclined to trust anything they say without carefully validating it.

they are not lying, apple was quite literally sued successfully in France and are currently settling class action lawsuits over the exact thing.

apple can say otherwise but that doesn't mean you need to regurgitate their lies. Apple only slowed the phones after a software update, not as a natural response to battery degradation. if that were the case it wouldn't be unique to that specific phone anyways.

God, apple makes good hardware but the people who just makes things up to stand in the fire of criticism are the strangest

More accurately, the French DGCCRF fined them for not notifying users that it was happening:

https://web.archive.org/web/20200207222719/https://www.econo...

The lawsuits currently being settled are the same: they haven’t stopped doing it, but now there’s a clear UI explaining why it’s happening when it happens.

> Apple only slowed the phones after a software update, not as a natural response to battery degradation. if that were the case it wouldn't be unique to that specific phone anyways.

It’s not unique to that model, or even Apple, as was widely covered at the time. They added that battery management feature in a software update in response to data showing a fair number of users were affected by unexplained shutdowns, so it affected the population of people who had been aware of their phones’ battery health but the problem was that it happened without notification instead of popping up a “your battery is dying!” warning.

You might find https://www.geekbench.com/blog/2017/12/iphone-performance-an... interesting:

> Users expect either full performance, or reduced performance with a notification that their phone is in low-power mode. This fix creates a third, unexpected state. While this state is created to mask a deficiency in battery power, users may believe that the slow down is due to CPU performance, instead of battery performance, which is triggering an Apple introduced CPU slow-down. This fix will also cause users to think, “my phone is slow so I should replace it” not, “my phone is slow so I should replace its battery”. This will likely feed into the “planned obsolescence” narrative.

You are being extremely disingenious (lying yourself?)

They are not "outright lying" like you claim, considering you explain why what they're saying is true ("phone throttling its power draw" = slowing down)

Purism is insinuating that the primary reason why Apple throttled phones was to force you onto a new one, which is false. I think most people would rather have a phone that runs slow than a phone that randomly shuts off (increasing the lifespan compared to doing nothing and making people think their phone was broken.) Where Apple went wrong was not telling anyone they were throttling their devices, and not letting you force them to run at full power (which they changed later after the media backlash.)

> primary reason why Apple throttled phones was to force you onto a new one, which is false

If you believe Apple's marketing, it's false. If you look at customers' reaction, it's likely true. There were a lot of complains after every update, and the only known solution was to buy a new phone.

Or replace the battery. I had one of the affected phones and it performed exactly the same with a new battery as it had with the original one during the first couple of years.

The whole point of the class action lawsuit is that most people didn't know this.

That’s not the claim which Purism made or which is being repeated in this thread – because there’s obviously limited marketing value in saying “battery health used to be hard to tell on our competitor’s device but it’s been easy for over 6 years”.

Reread my above comment. The effective result of Apple's action is forcing people to buy next phone.

Yes, that’s the claim Purism is using in their marketing but it’s predicated on the reader not checking the details and realizing that to the extent it was ever true, it was limited to a short period of time in 2017 between when the battery management behavior changed and when the UI warnings were added.

It’s also worth noting that multiple government investigations and lawsuits have failed to turn up any evidence supporting the conspiratorial claims about forced updates.

I know that Purism’s marketing strategy has to be convincing you to buy a product which is inferior on many counts but I think it’s a mistake to do things like this attacking their competitors because the “free is better” framing is fundamentally about trust. If they’re dishonest about something we can easily assess, how much can we trust the claims they’re making about things which are much harder to prove?

> it was limited to a short period of time in 2017 between when the battery management behavior changed and when the UI warnings were added

You are technically right, but important part you are missing is that the Apple's behavior has only ended after they were sued.

> things which are much harder to prove

Which things? All code is FLOSS btw.

> You are technically right, but important part you are missing is that the Apple's behavior has only ended after they were sued.

You have the order backwards: most of the lawsuits were filed after they’d shipped two rounds of UI indicating when battery health was degrading performance – the lawyers recruiting clients knew that would make it easier to argue that the company was effectively admitting fault.

https://www.patentlyapple.com/patently-apple/2018/01/with-tw...

> Which things? All code is FLOSS btw.

Ever wonder why they advertise “less binary blobs”? Any complex device has a long trust chain - you have the OS, of course, which we already know has an older security architecture than iOS or Android, but you also have the firmware for every component, implementation choices for how those components connect, and things like the CPU and it’s microcode. I guarantee you haven’t examined all of those so you’re trusting them to do so, which is what everyone does, and that’s why I mentioned the transitive nature of dishonest marketing. If they’re playing fast and loose in one area that makes it harder to say that they wouldn’t try to cover up something else, overstate the degree of diligence that they’ve applied, etc. If one of their developers is compromised, how do I know the same marketing weasel won’t decide that it’d be bad for their reputation to acknowledge it in the absence of proof that a signing key was leaked? I’d like to say that they wouldn’t but clearly their senior management aren’t placing enough emphasis on honesty.

Now keep reading all the way to the third sentence in the quote, where they characterize the motivation as getting people to buy new phones, when Apple, and other manufacturers, claim to be doing it to keep older phones usable.

Now, if it was some random internet poster I might excuse this as fanboys being tedious or spending too much time on clickbait sites but this is an official publication from a direct competitor. They have to have expertise in this subject to be capable of building devices and anyone capable of designing a phone is going to know this is a real problem.

> A single intrusive Android, Apple, or Microsoft app enables the developer to conduct audio, video, and physical surveillance on the end user while data mining highly confidential personal, business, medical, legal, employment, and location information from the end user to exploit for profits.

Y'all do know that modern Android and iOS have permission prompts to prevent this kind of thing, right? And at best this would still be "don't worry that our hardware is inferior because at least we make it harder to install malware" - deflecting doesn't convince me that speed isn't a problem.

> Y'all do know that modern Android and iOS have permission prompts to prevent this kind of thing, right?

You know these apps completely disable themselves if you don't grant them the permissions because the OS tells them when you do that, right?

That doesn't happen on iOS. Apps completely disabling themselves if not granted permissions is explicitly against Apple's terms IIRC. This is how I was able to install Truecaller (yes that invasive app) without giving it my contacts permission and then disable my number from it (which seems to be the only way). You can not do this on android as truecaller would refuse to move forward without permissions.

> You can not do this on android as truecaller would refuse to move forward without permissions.

Then it's only iOS that allows that. My point still stands though because the commenter said that "android and iOS" allow you to do this but they don't.

Back when Android allowed you to root the phone, XPrivacyLua let you spoof granted permissions so that apps thought they were tracking data that was actually fake. But now you can't do this. Some distributions like GrapheneOS and maybe LineageOS also let you do this nowadays, but it's still not enough.

I encounter this all the time with App Store apps. Citizen will disable itself if you don’t grant live, high precision location data even though it would work fine with “when open” and an approximate location. The app for sketchy WiFi cameras I bought requires location to register the cameras, and will refuse to activate them without the permission.

(Taking the claim at face value) Okay, then nothing is safe regardless of platform; apps on every platform can just demand permissions and be given them, and PureOS is no better because "we demand that you disable technical measures that would protect you" is effective on all OSs.

This still leaves the Librem 5 no better than anything else.

> This still leaves the Librem 5 no better than anything else.

It doesn't. The security model of GNU/Linux is that you run trusted FLOSS apps from a signed repository. Try to find malware in there.

Okay, then the security model of Android can be that you run trusted apps from a signed repository (say, f-droid). If you don't install apps that aren't from a publisher you trust, both OSs should be fine. If you do, then the only advantage of PureOS is that it's less likely to be targeted, not that it actually has better security.

Edit: As a worked example, let's say the app for some social network does something nasty. If they released their app for PureOS, it would still do something nasty. If they don't release such an app, then you can't use it, but you could also just not use it on Android and enjoy the same security.

Yes, sandboxing should be improved on GNU/Linux. Still, you can use the smart card on Librem 5 and kill switches if you need to be 100% sure malware has no access to some things.

These permission prompts is useless because way too many app asks for all these things.

There is still better security. On Android apps can't start using the microphone or camera while in the background. If an app is using the microphone or camera while in the background the app has to show itself on the notification bar. The notification bar will also have an indicator showing that you are being recorded.

Meanwhile on PureOS you can install a package and then it is free to spy on you all it wants without you even knowing about it.

This is not better security, it is wholesale: android and ios has security, while almost every linux distro lacks any form of meaningful security, which is a huge shame.

Linux phones have a different threat model. Saying they lack any form of security implies you know nothing about them: https://source.puri.sm/Librem5/community-wiki/-/wikis/Freque...

Do people not just say no? I reflexively deny any unexpected permission prompts these days.

And how can you verify that those permissions are behaving the way that you asked them to?

Google themselves have publicly stated that turning off GPS in your settings menu doesn't actually turn it off physically. So how do we know that apps still aren't accessing GPS data?

FWIW, I have been daily driving a Librem 5 for about 5 months now. It is NOT slow.

Some JS heavy pages take a long time to load (FF is mostly fine, GNOME Web is not) in the browser. Other apps work totally fine speed-wise.

There's a lot of opinions in this thread but not a lot of people with firsthand experience.

There's also a lot of opinions trying to be passed off as fact, e.g. saying Purism is lying when they said Apple intentionally slowed down their phones.

What do you do for apps like uber or other similar apps?

I have never used Uber/Lyft, so I cannot say.

There is one Android app I am using through Waydroid. It's a 2FA app I need to log in to work.

Uber/Lyft may also work through Waydroid, but I have no idea.

The phone itself looks like junk from 2011. The ghosting on the screen is apparent in the video demonstrating its "speed". The apps they are demoing look very poorly designed, but all of this to be forgiven because "they aren't collecting your data".

Disappointing but I didn't really expect much from them.

I can't believe this company is still around.

I don't think they will be around for long. Their products went from being the highly anticipated to most trashed for being totally non functional and hostile behaviour towards customers.

Luckily they have plenty of competition around: FairPhone, PinePhone, SHIFT [1] on the phone side and FrameWork, System76 on the laptop side.

[1] https://wiki.postmarketos.org/wiki/SHIFT_SHIFT6mq_(shift-axo...

> PinePhone

Not a competitor without a dedicated software developers team: https://news.ycombinator.com/item?id=37914312.

All others don't offer GNU/Linux phones.

Fairphone provides a version with a de-Googled Android (/e/OS) and I think that enough for people who care about their privacy and don't want to share their data with Big Tech.

Having an actual Linux phone with the same userspace as a Linux Desktop is cool, but it's even more niche.

It's definitely niche, but the exclusive features can be useful to quite a few people I believe: https://source.puri.sm/Librem5/community-wiki/-/wikis/Freque...

As much as I like pine, the pinephone is heavily subsidized by librem who actually hires software devs to make things work. If you like your pinephone, you have librem to thank for it in large parts.

For years linux people we asking for hardware like pine gives: just let us have it and we will write drivers and make it work. Now that someone delivers they get criticized for it. While Pine isn't perfect, they are filling a niche that I've long wanted.

I didn't criticize pine, I said they were being subsidized by someone paying for the software work, which is objectively true

Yeah they aren't fully equivalent offerings. But you could spend $800 on a donation to the KDE Plasma Mobile, Unity8, SXMO projects, $100-$200 on the hardware itself, and come out even, no?

Not everyone uses or wants Phosh/GNOME.

They've had a resurgence in the Linux Mobile community, now that they've caught up with their shipping backlog.

Librem 5 is the most suitable/polished/stable daily driver for Linux Mobile at the moment.

Purism are funding a lot of the development (Phosh, in particular) that other phones benefit from.

Purism's other products don't seem to have as strong an advantage over their competitors though.

> Librem 5 is the most suitable/polished/stable daily driver for Linux Mobile at the moment.

That's not saying much.

Please don't post shallow dismissals... A good critical comment teaches us something

https://news.ycombinator.com/newsguidelines.html

I am not sure in this case, is it a stable experience?

Linux mobile (non android) does not has the best reputation, which is what this comment is hinting at.

Working the best can mean 80%, when users expect 100% of basic functionality, with no regular restart and short battery life. So .. do you know details of the current state?

I'm using Librem 5 as a daily driver. The battery is sufficient for one day, but not more. The experience is definitely not 100% but maybe 95%+. Here is a list of complains, most of which I personally do not have (they are likely connected to older hardware): https://forums.puri.sm/t/l5-items-that-still-need-to-be-poli...

Apart from that, GNU/Linux provides many features that you can't get elsewhere: true convergence, desktop apps (including Firefox with all plugins), full terminal and so on. And no tracking unlike the duopoly: https://news.ycombinator.com/item?id=26639261. Also, hardware kill switches and a smartcard are really useful.

Hm, to be honest this sounds more like 80% to me, but is still in a state where I might consider one soonish.

Yes, from my link, it's more like 80%, but as I said, most listed problems I personally don't have.

Fair enough. Allow me to elaborate.

Being the best in an extremely niche market with little competition doesn't really mean very much. Purism is currently appealing to the overlap of people who specifically want a GNU/Linux (not Android) device and those for whom poor performance (compared to other flagship devices) is not a deal-breaker. Each of those segments of the market is already very small, and the overlap between them is minuscule.

The only way that a Linux phone could be economically viable is if it's rock solid and performs well, and based on the first-hand accounts that I've heard/read, Purism hasn't cracked that yet, which makes me pessimistic about them ever getting there.

> The only way that a Linux phone could be economically viable is if it's rock solid and performs well

This is where we disagree. Generally, this is not true, as proven by Pine64 business model. Of course the Librem 5's price is much higher, but it offers more features that could be useful to many: https://source.puri.sm/Librem5/community-wiki/-/wikis/Freque...

> Generally, this is not true, as proven by Pine64 business model.

Pine64 hasn't proven anything. I bought a PinePhone in excitement when it first came out, and after the first two weeks, it's been on my shelf collecting dust. I imagine that most buyers are in the same boat. Yes, they built up some hype and got some sales early on, but I think it's unlikely that this will continue into the future.

To be clear, "useful" and "economically viable" are not the same thing. It's great that these options exist, and as long as they're around, they will continue to be useful to some subset of people, but I think there is little chance of any of these Linux phones replacing my Android phone for daily use over the next 5 years—and I say that as someone who's pretty tolerant when it comes to bugs, poor performance, and lack of features.

I agree, it's not ready yet. What we see is a community forming, developing the software/protocols/skillsets needed to make Linux Mobile viable. It's expected that it's messy in the beginning. :)

> suitable/polished/stable

> Linux Mobile

Choose 1

Librem 5 is the most suitable/polished/stable daily driver for Linux Mobile at the moment.

I'm curious if some Linux phone can consistently make and receive calls, its battery lasts three days and I can buy it for less than 200€.

I only use calls, the camera, whatsapp (I could try a replacement), the alarm clock (a lot), sometimes vlc and sound recorder. Firefox, maps, etc. are nice to have, but not needed.

I wonder if I'm one of a kind user. But the three requirements doesn't seem much and support my current usage.

Pinephone is the only GNU/Linux phone that cost $200. Calls work, but I'm not using them much. The battery lasts one day, but you can have a spare one. Whatsapp installation can be complicated, since they (intentionally) don't have a Linux app. But some people I saw did install it. Firefox is the desktop version btw. The camera is very basic and slow.

Thank you! That looks promising. I would install some alternative app for messaging, but even whatsapp has a web version. I don't need much of a camera.

Calls work, but I'm not using them much.

For that I do need some reassurance :) I'll take a look to their forums to see how progress goes.

> I don't think they will be around for long.

I'be been reading such statements on the Internet since their first Librem 5 announcement in 2017, or even earlier.

Software wise they've done good work.

Speaking from trying alternative Linux devices eg. Pine64 I'm just done with it. I bought almost all of it eg. both phones, newest arm tablet, laptop, earphones... but yeah end of the day it's just a PITA. I'm done with managing how my ram works or not having suspend. Unfortunately I'm not a god dev like megous to bust out a camera driver.

Get an old flagship for same price and get better use.

Yeah I pulled out. I wanted it to happen but idk it sucks. I hope in the future there are more options than just Apple/Android.

Tangent. I have this "fetish" it seems of low-powered compute. even though I don't need to eg. I have an i9 desktop. But it was cool to dock a phone/use it on a computer... but I guess right now you have to pay for it eg. a Samsung note flagship (even old) so you have enough ram. But it's like intentionally crippling yourself for a fantasy use case (my mindset) eg. writing code while traveling in another country with your phone. "But it would be cool to do".

In the future IDX will be there, just login to a dumb terminal but give up your ability for cloud too.

The header bar with nothing but the app's name seems redundant and gives me the impression that I'm looking at a phone-sized window in GNOME rather than an app designed for the entire screen.

Actually, it is a phone-sized window in GNOME: https://news.ycombinator.com/item?id=19328085

Really hard to make this stuff. It would be pretty neat if whiteboxed phones would show up that are the equivalent of Clevo/Sager that other brands could specialize to and relabel. I imagine there's a lot of work in making the next phone in each of these cases that is probably better spent on the software platform.

Good for them for trying, of course. Impressive work.

Incredible. Purism has come a long way. It's amazing to have a truly open phone. A lot of people don't know how hard it is to mod an Android phone. You have to worry about drivers and firmware, security protections and attestation. For me the value of this comes from its freedom. Not from its privacy necessarily, but the hardware switches are great. I hate that I rely on the very many apps for Android and iOS that don't have a web counterpart like Discord. Linux apps on phones are very much a chicken-egg problem and I don't think we can break the duopoly

The video looks pretty good, time to dust off the Librem 5 in my closet and update its software.

Absolutely nothing about the demo video looks fast.

Maybe you're looking for "instant"? I just picked up a Pixel 8 Pro, and even that's not "instant".

The touch interactions look laggy from my perspective, but I could be wrong.

stop using and promoting linux phones

https://madaidans-insecurities.github.io/linux-phones.html

Thanks for providing a nice list of security drawbacks for Linux phones. Nice to have them all gathered in one place :)

However, I did not buy my Librem 5 to get a phone with feature parity with Android or iOS, I wanted a hackable phone that behaves more like my other devices running Linux and to support a future where users have the option to control and inspect their phones.

I'm following Fedoras efforts of making an immutable base system for the Librem 5 with great interest. I believe their approach will address at least some of your security concerns.

I use silverblue. It does nothing to address the concerns listed in the article.

Does Fedora/Silverblue have the same security problems (not applying exec-shield patch, disabling kexec etc) as PureOS does?

The article is somewhat poorly written despite making salient points. The core point is this part:

> There is not yet a single Linux phone with a sane security model. They do not have modern security features, such as full system MAC policies, verified boot, strong app sandboxing, modern exploit mitigations and so on, which modern Android phones already deploy.

The example about PureOS is that their hardening is minimal even for desktop linux, let alone for mobile. If you wanted to harden linux well enough to be on par with android, you would end up with android. That's the core point.

More info: https://madaidans-insecurities.github.io/android.html

Any mention of silverblue/immutability is not relevant here. It has nothing to do with the above concerns.

Reply on news.ycombinator.com