Hacker News new | past | comments | ask | show | jobs | submit login
PostmarketOS Revolutionizes Smartphone Hacking (drewdevault.com)
100 points by als0 on Nov 26, 2021 | hide | past | favorite | 34 comments



When looking back like that, it's mind boggling to realize how far we've come. I sincerely thank everybody who is mainlining their devices, reporting bugs, sending patches, reviewing and testing patches, maintaining packages, updating the wiki, helping out others. Without such a huge and skilled community, we wouldn't have made it this far.

And yes we have a challenging road ahead, as the blog post mentions. But at the same time, I believe the wider Linux Mobile community has reached critical mass of developers and projects. There are so many great projects in this space now (which collaborate quite a lot, Mobian developer kop316's MMS efforts is just one great example). Multiple phones where mainline support has a bright future. Even if a few of these projects or even phones should fail, Mobile Linux can still carry on. Let's find out where we all can carry this effort in 2022 and the years to come!

Thanks for this post, Drew!


I was always horrified by XDA solutions, and wondered how apparently so many people didn't find that scene incredibly sketchy and suspicious. And when more credible community open source models were already widely known.

I'd be very happy if PostmarketOS got more progress towards solid mainline kernel support for some devices -- including for the nascent 4G hard requirement of my carrier. (I tried to help a short while, but I can't currently. And I probably won't spend more time on it until I can be doing userland app and UI/UX development atop a solid foundation.)


> I probably won't spend more time on it until I can be doing userland app and UI/UX development atop a solid foundation.)

I wonder what you could do with GrapheneOS, maybe make it convergent, USB C dock?


I'm running GrapheneOS right now as an interim solution for my main phone. But the massive size of the Android source base, and the incomplete alignment with goals of who drives Android, means I'd really rather a solid mainline kernel (and X/Wayland and hardware platform take off. Then I suspect one motivated person can evolve a much cleaner and open userland environment foundation on top of that, while using some Linux desktop apps in the meantime.


Oh, most people do find it sketchy and suspicious. Hence they just stick with stock ROMs or solutions that a "big" and "reputable" company provides.


While cellular capabilities are restricted by binary blobs or missing firmware, currently postmarketOS is useful for repurposing old Android phones and tablets into single purpose devices, e.g. simple photo frame, weather display, radio player... Possibilities are endless.


I don't understand the point of this article. Why is PostmarketOS singled out over the 20 or so Linux distributions that run on the Pinephone?


Honestly, I don't think pmOS is any better than any other Linux on the Pinephone.

However, pmOS is pretty unique in their quest to mainline other, closed, Android smartphones by making shitty code vendors put out good enough to be merged and by making their own drivers and shims to get hardware to work.

I don't see why you'd need a Pinephone to work on pmOS, unless you want to develop a pmOS application or feature for an ARM device that has all the features up and running. It's more lightweight than some of its competition, for sure, but honestly it's not really better or worse in terms of hackability.


It runs on more than just the PinePhone, enabling reuse of older smartphones.


I love the ideas behind pinephone, but I think it's just not powerful enough. I'm not sure if there's enough access to more powerful mobile phone processors though.


The new Pro model has a significantly better processor. They're using a binned and underclocked 28nm Rockchip RK3399S 64bit SoC – 2x A72 and 4x A53 CPU cores @ 1.5GHz. The They used an A64 in the base model which was a 40nm Quad-Core ARM Cortex-A53.


I bought the pinebook pro and ya it just doesnt quite have the power I want it to, but on flipside the lack of power gives the giant battery tons of runtime.

But for the pinephone. I really want >= 32mp camera. Whereas 5MP is ridiculous and 2MP selfie? You're right it's slow, it's not quite figured out entirely.

Pinephone pro coming soon and also doesn't quite hit the wants for me. It does look fast though in specs.


A beefy camera would be very interesting to work with. As of now, our phones do a ton of proprietary processing which makes open source/external solutions... suck. I'd love to have a decent open source camera app with the capacities of something like GCam.


>A beefy camera would be very interesting to work with. As of now, our phones do a ton of proprietary processing which makes open source/external solutions... suck. I'd love to have a decent open source camera app with the capacities of something like GCam.

I'm a big fan of open source, literally have open source projects with many stars on github. However, I'm not one of those 'no proprietary blobs' folks.

Sony has a camera chip that's awesome, available, but comes with a blob? Fine with me.


Me neither, though if there's a decent and open source solution available, you know that's going on my devices ;)


PostmarkerOS currently support some Snapdragon 845 and Snapdragon 821 devices, so you're not limited by pinephone.


But AFAIK in this case, you're limited by the proprietary firmware tied to old Linux kernels.


No, 845 runs on mainline, my OnePlus 6 is on an 5.16 release candidate at the moment.


How does it work without free drivers? Will it run on all future kernels?


How's the experience so far? Does everything work?


Powerful enough for what? Get rid of regressive virtual machines and surveillance framework bloat and there's much less computation to be done. I'd love for the Free software mobile ecosystem the coalesce around lower power processors and resource-respecting languages.


The first iPhone had a slower SoC, yet it felt way faster. I think there is room for improvement on the software side to make things faster.


Librem 5 is significantly more powerful, despite the CPU is the same. Much faster RAM, GPU etc.


The Pinephone Pro coming soon is probably a way better deal than the Librem 5. Pine64 is also way more trustworthy than Purism, who is constantly lying about their products.


The Librem 5 is currently $1200 after multiple price hikes and looks to be significantly slower than the $400 PinePhone Pro.


Of course PinePhone is cheaper: Pine64 develops no software for it, the software developed by Purism is used by more than half of the PinePhone users.


> who is constantly lying about their products

This is a bold claim. They have a history of delays, but lying is not what they do.



They've been having a hard time with the funds, so they probably have to delay refunds in order not to go bankrupt. Recently they got some inestments, so it should get better: https://news.ycombinator.com/item?id=27949435. This is not exactly "constantly lying about their products".


They’ve been seeking investments from anyone on their mailing list for years. And if they can’t refund people for six months or more (contradicting the terms of their already insane return policies), how can anyone expect them to be in business long enough to ship out phones or laptops or anything else?

If you order something from Purism today, it'll probably be 6-12 months at the earliest before you get your product. Lead times on phones are over a year and there are plenty of reports from people who ordered more expensive Librem 5 USA models who still don’t have them after six months or more. Librem 14 laptops seem to be on a six month delay.

And yes, they do constantly lie about their products. They lied about FCC clearance. They lied about when the camera would work. They’ve lied/refused to address the reality that the cellular modem doesn't support VoLTE and likely never will (which is a problem because of how many US 3G towers are getting shut off and what that does to phones that can’t do VoLTE). They’ve lied about how many units they can ship. They’ve lied about the state of their software.

At least Pine64 knows how to ship a product to people and be honest about their device capabilities.


> it'll probably be 6-12 months at the earliest before you get your product.

Yes, they have a history of delays. By the way, did you heard about global supply chain problems?

> lied about FCC clearance.

I was following them for years. There was a delay in obtaining the clearance. But where did they lie about it?

> They lied about when the camera would work.

First, this is called “a delay”. Second, if you would follow what exactly happened, they had to reverse engineer the camera due to extremely poor documentation. And all efforts were in the open. Unprecedented transparency.

> They’ve lied/refused to address the reality that the cellular modem doesn't support VoLTE and likely never wil

They never lied about it. Any proofs? They have been working on it for a long time indeed for the same reason as above (documentation). The hardware is capable of 4G, so I’m sure it will work. Why do you think otherwise?

> They’ve lied about how many units they can ship.

They never disclosed how many units they shipped.

> They’ve lied about the state of their software.

This is the only smartphone company I know who develops all software in public. You are lying here.


I love to see validation from Drew of my impression ca 2014, trying to flash some ROMs onto a newly-unmaintained phone:

"Seriously, the solution to not getting security updates anymore is installing a binary using totallyTrustWorthyFoneFlasher_4 (Copy).exe?!"


My own preference would be for the pmOS project to pick a target subset of the devices on the Community and (perhaps) Testing lists. These would be chosen for a combination of criteria. For example (ducking):

- How close are their ports to being complete?

- Are they based on the aarch64 architecture?

- How many of them are economically available?

- How easy will it be to install pmOS on them?

With this subset defined, the project could focus on making these devices available to relatively naive users. There could be FAQs and HowTos for installation and setup, specific forum and wiki pages, etc. The knowledge gained in this sprint could help the project to develop an approach for the greater, long-haul effort.


After reading this post I decided to go buy an old Android phone and install pmOS on it.

But then I looked at this list. Is there really no device that supports a camera? This seems like an incredible omission if true. Am I misreading this list?

https://wiki.postmarketos.org/wiki/Devices

Is there a good device for another distribution that at least has calls/texts, camera and network support so I can browse the web?




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

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

Search: