Librem 5 Evergreen vs. Pinephone
thatgeoguy.ca
thatgeoguy.ca
> Linux phones are definitely not yet ready to be daily drivers
Despite the author making this point twice and some of the negativity regarding Purism's pricing - I still feel very good about the state of Linux phones and the incredible work these two entities are doing. We can't lose sight of this.
It wasn't too long ago that the mere mention of a Linux phone was pure fantasy. We now have two companies (and two passionate communities) that are helping move Open Source Linux forward. I'm still very excited about how things are developing and the trickle down effect it will have as we move forward.
> One of the neat things about the Pinephone is the ability to boot and use a variety of Linux distributions on the hardware itself. It’s as easy as loading the distro onto an SD card, and then booting the phone. You don’t even have to flash anything to the internal eMMC in order to run it!
This is such a killer feature for me - while also providing a modern-ish platform for the various distributions and app developers to test their wares. It's probably the biggest single most important thing about the PinePhone right now.
No more unsuccessful rooting, flashing and possibly bricking. No more nagging feeling that you are merely renting a device and are constantly at the whim of a corporate overlord (which never ends well).
I think, like the author, I'm a bit more excited about the flexibility of the PinePhone but acknowledge the importance of the work Purism is doing to try and bring a First Tier Linux phone to market. We all benefit from this.
We’ve come a long way since the OpenMoko :)
FWIW while the Librem 5 can't run the bootloader straight from the SD card, the kernel can use the rootfs from SD card just fine - I have multiple distros installed on SD card there this way.
The Librem 5 also supports booting via fastboot (uuu), which personally I find to be a more convenient option for emergency fallbacks than having to juggle with SD cards like you have to do on the PinePhone.
FYI - the reason why L5 doesn't boot straight from the SD card is that the SD card reader is on the USB interface behind a hub, because both SDIO interfaces available on i.MX8MQ are already taken (one by eMMC and other by WiFi).
Shame though, because among all the Pine devices I have (Pinebook Pro, Pinephone) I've gotten very used to booting from the SD card to swap things out.
Have you written / do you have any links to documentation for how to use fastboot?
Some suggestions, as a user, to get it over the edge:
1. Every app in the Manjaro distribution that ships pre-installed should at least fit on the screen; many apps have popups that do not.
2. Dialing should never cause a reboot; calling voicemail has crashed the phone for me.
3. Either give up on having a camera or make it _fast_.
4. Some kind of loading indicator for apps would be nice.
5. Either don't allow X apps to be installed or make the X wrapper work; I can install gtk emacs but it doesn't run.
Sounds like an issue you should bring up with the application authors. These are usually easy fixes.
> Either give up on having a camera or make it _fast_.
We've already seen demos of the camera running at 60fps. Why would you want people to "give up" on a perfectly achievable goal?
The application authors are mostly unaware. The Manjaro ARM team is using libhandy to port desktop applications on a case-by-case.
They are also in charge of what applications are preinstalled; IMHO, they shouldn't be preinstalling any applications which do not _at least_ fit on the display in every popup and window they open.
> We've already seen demos of the camera running at 60fps.
It displays at 60fps, but I still can't seem to get it to take non-blurry photos of my kids.
Sure, I could have lugged my full-size camera along and then sent it over to the phone via wi-fi, but using a real camera in a supermarket is going to get you some unpleasant attention from staff.
That's a bit of an understatement. It's not 'a lot'. Everything running on those phones is a GNU/linux application. The fact that there are binary blobs and that the UI was closed source doesn't take away from that. The post I was replying to was complaining about linux phones, not about "open hardware/open source" phones.
1) managing 3G/2G network. There seems to be quite a few hacks in the telecom infrastructure and I have long suspected that mobile phone operating systems are simply expected to handle each unique occurrence. I expect calls will occasionally fail, especially for answering and calling, but also in transitioning between networks.
2) Bluetooth audio support for calling is already known to be broken. I can only hope it will be fixed some time in the future, but I suspect different devices for handing handsfree calls will have their own hacks which won't be supported.
3) As the article alluded to, the GPS software has a bunch of hacks to address short comings. I am expecting there to be both long delays at finding signals and strange jumps in accuracy. Since my personal interaction with a GPS tend to be in moments where delays will be very noticeable, like when you entered the car and waiting for the map to show where to go before starting the car, delays and inaccuracy will be noticeable in such use.
4) For music it will need to interact with other devices, especially over Bluetooth, and I strongly suspect each manufacturer has it own hacks in how to do track progression, rewind/forward, play lists and so on.
5) Android app support. Here I have some hope that Anbox will work. I am however worried that a lot of apps will break in strange ways when run in a emulator from missing dependencies or strange hacks.
Also, while Anbox has been used in some cases so hackers can say "See, I did it", I don't think Pinephone users are actually going to be using Anbox on an everyday basis. It is too RAM-heavy and too taxing on the Pinephone’s weak processor. People expect to be able to near-instantly switch between apps (e.g. between Signal and Firefox), and that is probably never going to be possible on this iteration of the Pinephone.
With regard to music and Bluetooth, however, already Mobian on the Pinephone gives me the same functionality with my Bluetooth amp as my Android phone. No complaints there.
> it’s kind of appalling that literally anyone would pay this much. It’s kind of appalling that Purism would charge this much. I can imagine a scheme where the parts are shipped from China, and verified and assembled by a part-time college student in the States and they’d likely still be able to price it cheaper than $1999.00.
There are definitely valid reasons not to buy Chinese products: https://en.wikipedia.org/wiki/Boycott_Chinese_products. Also what are the other competing smartphones which are made in USA?
We impose labor and environmental rules on products made in the US but don't require the same standards for products made overseas, even though we could.
China is basically a trust that uses profits in industries it has already taken over to subsidize its entry into industries it hasn't yet, and for decades we did nothing so we could get cheaper shit, claiming that we would keep making the high tech stuff, until we turn around and every phone and computer is made in China.
There are solutions to this. Obvious solutions. But it would imply paying more for stuff because it's not being made in a polluting sweatshop in an undemocratic country which is acting like the nation state version of a 19th century robber baron. (At least initially; you know what stage two of the robber baron playbook is if we do nothing.)
Well, even if you do, you are up to myriad of other issues.
The progressing general business unfriendliness of the West will not go anywhere. If it does not go, the situation will not change even if you put 1000000% tariff on Chinese goods.
I myself been working on setting up a small knock down kit assembly factory 6 years ago in Richmond, Canada. It was a nightmare. We also attempted going across the border to Washington with same results.
If doing even something as simple is that is a guaranteed to fail endeavour, than, what else you think you can do?
The Ease of Doing Business index is a complete BS.
Western countries in their majority should not be far away from countries they scorn as "third world" for how hard it is to run a business.
If you calculate costs purely on paper, running a factory in South China is already more expensive than in USA. It was so for quite some time.
You do not see Chinese factory moving to US in droves, despite some early optimism.
Chinese entrepreneurs saw for themselves that behind attractive land, utility, services, and some times labour costs, hides an impassible bog.
Should we allow a factory in Indiana to avoid environmental regulation if they just pay a 30% tariff? Is it ok to have a 5% mortality rate at a factory in Alabama if they pay a 60% tariff? Should U.S. companies be allowed to just pay fees to become exempt from FDA/EPA/OSHA/etc. rules?
I know some will argue that it is the case in the U.S. A mining company can break the rules, kill a dozen workers, then pay a million dollar fine and go back to business as usual.
But we don't investigate, we don't fine, we don't punish in any way foreign companies that don't follow U.S. environment/labor/safety rules. We open the door and put their products on the shelf right next to U.S. products. The answer is not to levy a tariff to make them price-neutral. The answer is to ban them just like we should ban U.S.-made products that don't follow the rules.
If a company can pass inspection and certify that they meet U.S. rules the same as U.S. manufacturers, and still have a price advantage, I say let them clobber U.S. competition. But the U.S. company gets regulator inspections as part of their taxes. Foreign companies have to pay for U.S.-recognized inspectors on their own dime.
Same thing for U.S. soy farmers selling into China or Norway. They should have to meet the standards of those countries and pay for the inspections that prove it.
Even now I know factories that only export to the EU, or only ship to America because supporting 2 sets of complex regulatory compliance requirements isn't manageable given their team size.
There are problems that tariffs are a solution to, like currency manipulation. Lack of environmental rules is not really one of those problems.
> The answer is to ban them just like we should ban U.S.-made products that don't follow the rules.
The real problem is that there are two problems, which currently sort of cancel out.
The first problem is that the existing rules in the US are effective but prohibitively burdensome. The second problem is that the same rules don't apply to China. Businesses solve the first problem by invoking the second.
We can't solve just one. If you imposed the existing rules on China, things would become unreasonably expensive. If you stopped imposing them on the US, there would be as much pollution in the US as there is in China. We need more efficient rules. Which is hard. Worse, it's not sexy and nobody wants to do the work.
But in a sense we can have one solve the other in the opposite direction too. Apply the same rules to other countries and they can't be avoided by leaving the US, which then creates pressure to fix the rules so that they're not prohibitively inefficient, since they can no longer be avoided.
For a time, yes. But the large market reachable by meeting those requirements would lead to innovation in trying to meet those requirements. (Or in trying to cheat those requirements, which would require some care to enforce.)
They can also be enforced via a ratchet upwards, so that there's time to adapt.
Is this an argument to repeal the XIII Amendment in the US?
People keep talking about environmental regulation, but I'd say that they alone do not account even for a few percents of all problems hampering US manufacturing.
Same is for labour rates, US labour rates in non-flyover states are already lower than in South China...
Seeing people keep talking about that, and fail to see way bigger elephants in the room.
Eh aren't they basically non existent? Like 6.2% unionised in private industry and that tends to be more in other industries.
I find it hard to blame them for this later migration. Hell if you'd take their lobbying on foreign policy they'd probably be fairly protectionist.
At the end of the day I think the major dealbreaker is that everything in china whether it's low skill labour or high skill labour like engineers automating things is cheaper and that the capabilities exist. (Some part of bumfuck Africa or SEA might also be cheaper but not have these capabilities) There's also plenty of competition there driving those prices down vs...well what's left in the US and EU.
That wasn't the case when this started. The decline of unions in the US and the movement of US manufacturing to China has pretty well gone hand in hand.
> Hell if you'd take their lobbying on foreign policy they'd probably be fairly protectionist.
Protectionism apropos of nothing is a non-starter, because there's no argument for it. You can say we need tariffs on China to account for currency manipulation and prevent a single undemocratic nation from monopolizing important world markets, but you can't win an argument that says food workers should pay more for transportation just to benefit auto workers who already make more money than they do.
But they do often win arguments about not automating away jobs, because there is less money to be made automating them than there is in getting paid to do them when they're not automated, since the automation would apply to everyone in the industry and just lower prices. So we've been fighting automation for the sake of jobs even though it's the main thing that can make higher priced US labor competitive with lower priced foreign labor, and then we lose the jobs anyway.
Do you have an example where this happened?
Short-termism around corporate earnings has done more damage to U.S. industry than unions ever could. And to be clear, I don't think China is any better, despite popular (misconceived) impressions to the contrary.
> But it would imply paying more for stuff because it's not being made in a polluting sweatshop in an undemocratic country which is acting like the nation state version of a 19th century robber baron.
That's increasingly not applicable with automation as well as the plunging cost of renewable energy. There's natural gas being flared in the Dakotas because it is uneconomical to deliver it. And as I understand it both North and South Dakota are not known for an overly restrictive regulatory environment or leadership. It's more a matter of volition IMO.
I've seen so many go through that cycle, and if they aren't involved in an industry with strong reasons other than money to manufacture locally it invariably wastes time and talent, and often kills the company.
Made in USA is a big stretch when nearly all components are imported from abroad.
Better to say "Few screws inserted in USA"
On why you cannot really make thing outside of China:
The biggest proponents of very purposefully locking manufacturing in China are major American companies...
I heard from multiple first hand accounts if Apple trying really hard to move out of China, and putting a concerted multi-year strategy of moving suppliers to TW, and Vietnam. If they cannot move, who will?
The emphasis on the later is because there is positive precedent of any major company succeeding in moving, and that validates beancounter opinion "If multinational A doesn't move, why should we?"
I'd also add that just because you don't want to buy Chinese products != Buy USA made products. There are a vast number of other countries that would jump at the notion of becoming your supply chain, and don't necessarily have all the ethical grievances of supporting China.
I actually think my main objection though had nothing to do with "buying Chinese goods" vs. "not buying Chinese goods." Mainly: why would you buy a $2k smartphone that is grossly overpriced and isn't really providing much more than the base model? As the above comment here says, it should really be closer to "few screws inserted in USA" than "Made in USA."
It seems fine as a statement piece if you're a big company willing to put your wallet on the line, it means a lot less if you're Purism and you need to stay focused in order to do it. And I will argue again: you could probably do this for vastly cheaper than $1999 even in the States. If anything it hurts this more, since people will look at this as an example and say, "Well, I'm not paying $2000 for something like that, guess USA manufacturing is dead since they can't produce good hardware at reasonable prices."
Given that no one currently is and that the average salary in the US appears to be about double the average in China - that seems about right?
An experienced factory worker (5-7+ years) with professional training, certifications, and clean track record costs at least 15k CNY a month in salary, and 6-10k more in pension/insurance/other social payments.
Factory workers in US get how much now?
> An early career Factory Worker with 1-4 years of experience earns an average total compensation of $13.28 based on 228 salaries. A mid-career Factory Worker with 5-9 years of experience earns an average total compensation of $14.17 based on 91 salaries.
A line manager working near a top city might get that... that's about the same as an engineer with a couple years experience makes. A high level engineer might expect 30k per month. Foxconn pays about 12k per month for non-technical managers with 3-5 years experience.
If you are a technician who assembles parts you are probably getting a salary between 3500 - 8000 RMB per month (+ social insurance payments if you work for a nicer/bigger employer). Maybe as much as 12,000 if you have advanced skills, but usually at that point you would be managing people as well.
Chinese, from my own knowledge of contract manufacturing. While saying this, I mean that it is not necessarily that all of workers are hired at such seniority level, and, yes, most of them will be seniormost workers on their line.
Foxconn, Flextronics, Pegatron — those companies are completely out of this league, they focus on the bottom line only, and do not work in markets where professional staff on the line matters.
One of our own manufacturing contractors who specialises on things including automotive assemblies for Germans (they did some of the new S-Klasse electronics,) and medical devices has such workers, and they do go through all the problem retaining them.
Out of other high end contract manufacturers, GoerTek had particularly good prices, and quality level despite them being somewhat volume oriented, but shame, they dropped all non-audio business as of late. Same was for LuxShare, after they got Apple's business, it became impossible to get through to them.
Some companies with own manufacturing also been known for well paid line workforce, BuBuGao, and Huawei instantly come to mind.
Frankly, ThatGeoGuy's assessment of the Librem 5 USA is ridiculous. First of all, he doesn't evaluate what it costs to set up a facility in a San Diego suburb that stores parts with supply chain security, orders small-scale PCBA production, and does small-scale final assembly. Second, he ignores the fact that the Librem 5 USA is a niche product for corporate and government clients who need supply chain security. It is competing with products like the Silent Circle's Silent Phone, Motorola Solutions' LEX L11 and Bittium's Tough Mobile 2. The only one of those phones which publicly lists its price is the Bittium Tough Mobile 2, which costs €1550 (US$1816), so the Librem 5 USA is in the same price range. Third, he totally ignores the cost of paying roughly a dozen developers to work for 3.3 years to develop a Linux phone. How does a company recover those high development costs that have gone way over budget? One way is to develop a side product with high profit margins that caters to niche customers with specialized needs.
This is simply wrong: https://puri.sm/products/librem-5-usa/. Mainboard PCB, Mainboard PCBA, Electronics, Modem PCB etc. are produced in USA. All Purism devices are assembled in USA, but this one is actually made in USA.
At least it does provide one data point on what it costs to do though.
Another obstacle is 3d printed case, phones need both high precision and high strength, AFAIK many 3d printers/materials lack at least one of these properties.
Still it looks doable, and apparently some people actually did similar stuff.
http://alumni.media.mit.edu/~mellis/cellphone/index.html
https://diy3dprinting.blogspot.com/2013/09/diy-mobile-phone-...
If the goal is to make just a single phone, and you don’t already own all that stuff, I’m afraid the required equipment is going to be above $1000. Maybe if you live in China will be cheaper, but not for the rest of the world.
For me, the "goal" of mainline GNU/Linux smartphones is that we get like PCs. I prefer fedora, you prefer Debian, they prefer Slackware, all is well, we are all free to choose whatever we want for our computer.
Here I'm reading it's all fragmented? And if it's fragmented, I guess it means it doesn't rely enough on Linux drivers, so long term support won't be there?
Is someone "wrong"? Like Librem is saying they are going mainline but actually aren't? Or is it Pinephone? Or I'm misrepreaentating what mainline should mean?
See https://debconf20.debconf.org/talks/13-my-phone-runs-debian-... (around 20:30)
I don't see any lack of contribution to the mainline kernel or upstream code from either community. The main divergence between the two is the price point and how the communities are run. As for fragmentation, well, if you want a choice in distro, then fragmentation does kind of happen as a result of that.
But overall a lot of the momentum is in the same direction. The biggest sticking point I see happened years ago, when Purism opted to make Phosh + PureOS and make it GNOME based instead of rallying more effort to the UBPorts project. That said, it's free software, so this happens all the time.
(Don't get me wrong though, I really love what Ubuntu Touch has become, and the UBPorts community too!)
Power management is complex these days. You'll often get a minimal implementation from the chip vendor, and it is a fair amount of work to really polish the lower level support. On top of that, you have to integrate this with the rest of the OS, so that active apps can increase the clocks, which then get automatically scaled back as system load decreases. So there's a lot to do on the userspace side as well.
Reading through the explanation, I think the tradeoff is still worth it for me since I care more about battery life and heat than full convergence (I plan to use this device as a portable Linux machine on the go, not so much (if at all) docked). I do think Purism should have been clearer about this though.
If you still want to do it, check this: https://forums.puri.sm/t/comparing-specs-of-upcoming-linux-p....
But I honestly expected few major phone manufacturers to join these development effort in some manner after seeing the fate of Huawei with Android or specifically Google Play Services/OHA(Open Handset Alliance), I'm surprised it didn't happen yet or perhaps it's because of OHA it's not happening.
Tizen didn't work out for smartphones although it seems to be gaining market share in wearables, why not direct the smartphone investment part of it to PostmarketOS, UBPorts etc. I presume Linux Foundation has to play a role in it.
Ironically it used to be a research project demoed a few times at FOSDEN and US gave them a reason to bring it into production.
Tizen is also being used in Samsung's smart TVs.
LG has WebOS and Sharp Aquos.
And then there is ChromeOS as well.
Basically, GNU/Linux desktop has proven to not be something that general public even considers when it comes with strings attached about FOSS politics and little UI/UX that those people actually care about, without the apps all their friends use.
EDIT: Typos.
Mainline GNU/Linux isn't something that makes average consumer pick vendor A or B when looking for a new device at the shopping mall.
And the hardware alone also doesn't sell, it must be the whole vertical experience on how software + hardware work together to bring that experience into existence.
This is exactly why Librem 5 is more promising as Purism develop both hardware and software.
Yeah, sure, the Librem5/Pinephone boot directly into open source firmware instead of some Qualcomm monstrosity that even acts as a hypervisor, but. Near-mainline Linux with GPU support on actually powerful devices with 3-ish-GHz fast cores, 8GiB RAM, and so on. And you can buy these on the used market rather than investing in more new e-waste.
If people on qualcomm phones had any amount of faith the company wouldn't screw them over or the community would be able to make a huge leap in terms of compatibility they probably would keep the phones but that's not how things have worked for about a decade.
It's more or less fine, especially considering that suspend is not yet implemented at all: https://puri.sm/posts/librem-5-4500mah-battery-upgrade/.
Eventually once the software is more refined, there are definitely some features worth adopting for. The kill-switches are a decent feature, and I would love to have something like that on every phone (especially the cam & mic one, since I don't tend to put tape over my phone cameras!).
For the Pinephone, the hope is that one day you'll be able to buy one (or some successor) and be able to load whatever distro you want on it. It's vastly easier to work with than installing a custom bootloader, and then flashing a new version of Android. So as far as that goes, there's a lot to be desired.
You also get the same software running as you do on desktop Linux - just hook up a display and you've got a regular Linux desktop. You can print from the device itself, because of course it can just run CUPS. You can direct audio output from each application (including to the earpiece) because it's just running pulseaudio. Hook up a webcam, a Wii mote, once you have actual control of the hardware the possibilities are endless.
(edit: rewording)
I want a pocketable Linux device I can take with me everywhere. That's it. I don't care how well it works as a phone. I want a fully-featured, honest-to-god GNU/Linux terminal in my pocket that I can use to do anything that I could do on a (massively underpowered) Linux laptop, especially on the command line (i.e. I don't even care too much about GUI apps except some basics like web and PDF display).
If it can someday replace my phone, that's a massive bonus of course, but realistically I don't think it will within the next decade, for my specific smartphone needs. But that's OK - I don't mind carrying alongside it an Android device that is my actual "smartphone".
Finally, you're not stuck with a PalmOS style mobile GUI and can use a normal overlapping window manager.
All this to say: you can't always serve everyone.
Judging by many, many reviews of smartphones, most people do deeply care about the quality and features of their phone's camera, and it's very likely to be the only camera many people have.
I *did not* / will *not* test:
Phone calls & SMS / MMS functionality
I haven’t bothered plugging them into a display.
Email (...) I’ve been too lazy to try setting this up
Cameras
GPS / location
(Article is not empty, does mentions how these machines behave on the GUI side, data-only SIM, web browsing, music, charging.)
- Kill Switches
- Charging the battery
- Temperature
- Phone shape
The rest of the article is not about Librem 5 vs Pinephone. While the sections on the apps could have compared the two phones, instead they compare two operating systems. I would have loved a discussion of the performance differences between the two phones or the screen brightness or the speakers. However, most of this article is largely irrelevant to the phones.
I have to say that hardware-wise, the Pinephone got very _very_ far already. It's an amazing hacker's device, especially with the USB-C host.
The only issue that persists is software. Phosh is pretty much useless because not a single app will work as a real mobile app, let alone act in a responsive manner with their underlying design philosophies. It all feels like Android 1.x, which was barely useable outside the context of shitty car navigation handhelds.
The available cross-compiled versions of Browsers (e.g. Firefox or Chromium) are totally broken, as they are made for Desktop and do not even downscale their user interface, and are neither responsive anyhow. So for now, I am using my own (prototypical) Browser Stealth [1] due to lack for alternatives because there at least the UI doesn't crash every few seconds when being used within a WebKit2GTK webview.
I don't have many requirements for my phone, to be honest. I don't use social apps, and only communicate via telegram. But currently the amount of software that's available for phosh is still the one reminding me of a developer environment or say, an Android emulator image.
I seriously thought about taking ownership in creating a few apps for it, like one for openstreetmaps, notes, text editor, which are the more important ones for me.
But to be bluntly honest, I don't even know where to start with the whole GTK shitshow. It's absolutely impossible to develop an app without any working examples. There's no non-auto-transpiled-bindings-documentation, there's no libhandy examples that actually work, and there's certainly not any responsive GTK app examples that would show how to make a "real mobile" app. How do people develop apps for mobile GTK?
Seriously, I cannot understand it. I want to, but there's no way anyone can keep the mood up digging through a _mobile phone's gdb output while the on screen keyboard crashed_ to figure out how the API works.
I think what GTK really needs is something like Electron, but maybe based on WebKit2GTK and (maybe?) gjs. Something that allows bundling your assets to build an app that is offline-ready, and is easy to build with common methodologies of UI frameworks.
Because currently I don't see the reason why libhandy exists. It tries to fix a problem, and doesn't fix the underlying cause. GTK boxes and their layouting (and reflow?) concept just isn't made for an adaptive/responsive UI, so you'll always end up with a totally broken UX by modern standards. (You're welcome to challenge me on this, but as long as "make a sidebar that can fade-in/fade-out with a swipe" takes more than a week to build, I think you're wrong.)
BTW: May I add your Browser to https://linmobapps.frama.io?
> BTW: May I add your Browser to https://linmobapps.frama.it?
Sure, of course, but as I said it's still very prototypical and needs a lot of work :)
PS: Really nice blog you got there! How's your experience with Lorimi so far?
Thanks! I have been busy with my day job lately, so I have just been trying to use my PinePhone with Phosh mostly. Lomiri is nice, but Manjaro Lomiri is not yet at the point where it does not break on upgrades, which makes it nice to play with but impossible to use.
I think many in the community are interested in this, but have been waiting for hardware to show up.
I wouldn't compromise my security by choosing one of this smartphones.
I had to manually search through the Google image to find what libraries existed, manually extract them, add them to my AOSP build, THEN compile. I was only able to do this because others had done much of the grunt work already (i.e. GrapheneOS), but when it was "upgraded" to Android 10, it all changed and all my work went out the window. Now I just use GrapheneOS and accept the comprimises I have to deal with.
You can buy the used devices cheaper than Pinephone and play with them. Not perfect hardware support, but enough for hackable mobile device with privacy and low-distraction mode.
>with privacy
There is no privacy with all proprietary drivers. There is no security either.
What will get me to buy a product like Librem or Pinephone in the future would be as a universal device, that is a dock that would make it my only digital device - and feel secure against hacks, etc.
I know that some Google provided APIs will be missing, but still, we could soon have alternatives, like OpenStreet for maps.
Ha, I can't help but think this is something linux folks won't even think twice about. :)
Also $1999 and the camera doesn't work yet? Yikes.