[skymer]

At best, the Pinephone is buggy software on low priced but mediocre hardware. It has always been like this, and it is not going to change. There is nobody willing to spend the millions of dollars it would take to make it otherwise.

Meanwhile, AOSP is fully working, open source Linux-based phone software that runs on lots of modern hardware from a variety of manufacturers. Distributions like GrapheneOS have put huge effort into de-googling, privacy and security. Cameras work. SMS works. Phone calls work. There are lots of apps. I hear there are even a few phones that can do wireless charging!

Why waste any more mindshare on the Pinephone?

[dbeley]

AOSP is still a google-controlled project that requires a ton of blobs, whereas the Pinephone (and other linux compatible smartphones like the Librem 5) has an entire ecosystem of healthy linux distros (from Mobian to Archlinux) running as close as possible to mainline linux.

> GrapheneOS have put huge effort into de-googling

They are doing great work but they're still depending on google for maintaining Android (Can GrapheneOS do anything against a feature pushed by Google in Android 13/14 and so on? Nope).

> Cameras work

You still lose all the post-processing stuff which account for a good portion of perceived picture quality.

[zozbot234]

The blobs are required by hardware OEMs, not by google themselves. You'd have the exact same issue when trying to run pmOS on the same hardware; it actually crops up throughout the embedded ecosystem. Pinephone and Librem 5 are quite exceptional wrt. minimizing the amounts of hardware-specific blobs that they require.

[dbeley]

Yeah indeed I should have specified that the whole "mainline linux compatible" is mostly because the Pinephone and the Librem don't use Qualcomm (or worse Mediatek) SoCs.

[MartijnBraam]

With a bit of work a lot of this can be improved, Android likes to have userspace drivers while a bunch of that stuff is also already in the kernel, so postmarketOS uses kernel drivers.

[kop316]

> AOSP is fully working, open source Linux-based phone software

Uhh....no.I know for a Pixel 3a, you cannot even boot AOSP without compiling in the driver binaries:

https://source.android.com/setup/build/downloading#downloadi...

https://developers.google.com/android/drivers

And there is an entire vendor partition (On the Pixel 3a, it was >400 MB). However, even if a user built AOSP with those drivers, VoLTE, SMS on LTE, Wi-Fi Calling, eSIM, and a few other things do not work.

To get those features, one has to extract them from a stock image:

https://wiki.lineageos.org/extracting_blobs_from_zips

I can't imagine this is any better for newer Pixel phones, nor for any non-pixel phones. This will be true for compiling AOSP or any Android ROM.

On top of which, that Pixel 3a is effectively frozen at the kernel it has when it reaches EoL

[betwixthewires]

Try changing the default DNS in AOSP from 8.8.8.8 without doing ill advised, hacky weird nonsense. Have you tried to prevent your AOSP phone from reporting tethering to your carrier in the last ~6 years? The software is licensed GPL, which means you own it outright as a user, to it's core, right? So these types of things should be possible with ease with a little technical proficiency, right?

Android, including AOSP, is not an operating system, because it doesn't enable the user to operate the device. It is a kiosk, because it enables the licensee to use the device, and limits utility to OEM, carrier and Google market requirements and business cases.

[aquaduck]

Well said. I'll have to remember the kiosk metaphor for Android.

[selfhoster11]

> Why waste any more mindshare on the Pinephone?

Android is a horrible platform to hack on. I'd much rather hack on something closer to base Linux - think Maemo or similar.

[megous]

> Why waste any more mindshare on the Pinephone?

This has some weird amount of assumptions in it.

People may have any number of uses for pinephone HW. It's a nice SBC with display, touchscreen, battery, a lot of connectivity, very efficient suspend to ram implementation (find me a SBC that will suspend and run at 60mW), I2C and UART connectivity and a very nice GUI bootloader available for really pleasant SW development at any level.

Why waste any effort writing apps for Android, when you can only run them on Android? SW I wrote for Pinephone, I also use on my other Linux tablet unrelated to pine64, and on a Pocketbook E-reader, and on my workstation, without any weird hacks or emulation. It's a nice device/phone for developers used to GNU/Linux way of distributing software. If you don't want to, there's no "security" getting in the way of doing pretty much anything with the phone.

To devs who did nothing with Android, it's a massive waste of time learning all the idiosyncracies of Android, just to be able to make some simple app, or deal with restrictions like no audio recording of calls, no playback of audio into calls, background tasks limitations, etc. If I want to run rdbms on my phone to store data, I'll have it running in 1 minute. If I want to do it on Android and connect to it freely from whatever app I want, it will likely take longer if at all it would be possible.

"Why waste any more mindshare on the Pinephone?" just shows a complete lack of imagination

It's a versatile device that can be used for a lot of things. I can pop my Pinephone out of a backcover and pop it into this: https://megous.com/dl/tmp/1551fe95bdfade31.png and have fun making HW projects with it, for example.

[ocdtrekkie]

On the contrary, AOSP is a largely nonexistent OS thrown over the wall by an evil adtech company so they can claim the proprietary malware-laden version they ship to 99.9% of consumers is open source.

You cannot remove the stench of Google from Android and we should stop trying. We should use a scorched earth policy of avoiding any code that Google has written and assume it is harmful by default.

[zozbot234]

AOSP has way too much churn for a sensible system. A Linux phone should just work, and share as much of its code as possible with Linux systems running on other device classes; distributions like pmOS and Mobian (and quite possibly Debian Mobile in the future) are working towards this goal.

[mindslight]

On my laptops/desktops/servers, I can rebuild all of my many systems from source, with config kept in git, with a few commands using Nix, and it basically Just Works. Before using Nix, I could reproducibly reinstall my personal Debian environment in 15 minutes with similar uniformly-managed config. When I upgrade software, there's no wondering if I'm going to be forced to upgrade hardware. I'll routinely keep machines running for 10-15 years with no downsides apart from higher power consumption. And when I do upgrade hardware, I'm never forced to change my software environment because of incompatibility.

Meanwhile installing and administering Android is an unreproducible dumpster fire in the vein of of MS Windows - just with tedious swiping instead of tedious clicking. Even my Lineageos/microG phone, with only Free apps, was never really the comfortable experience I expect from a personal computer. Rather it was just something "good enough" that I had to use when mobile. Its camera produced terrible quality photos, as it lacked the proprietary blobs. Even the best-in-class Google-manufactured phones running community distributions with no compunctions about proprietary blobs seem to produce subpar photos. And I'm still not really sure how self-administered Android updates are supposed to work, besides occasionally paving over the whole thing and then going through the pain of manually setting it up again.

That microG phone has since fallen by the wayside due to the 4G deprecation. I've presently regressed to a stock "full take" Android distribution on a no-charge "upgrade" that was sent to me for the 4G deprecation. There are no community distributions for this model, never mind degoogled ones, because of model churn meant to confuse the market. And I don't see the point to throwing $500+ towards a new Pixel (seemingly what all the innovative community distributions want), and rewarding Google for creating this technological pox where 3-5 years of updates is touted as if it's a long time.

A true Linux-first phone is putting a line in the sand of what Freely works today, rather than constantly playing catchup on the shifting sands of what the ewaste-surveillance industry is pushing this quarter. I'm sick of playing catchup, I'm sick of having to choose one failure from (commercial surveillance, no security fixes, upgrade treadmill). The next device I buy to carry around in my pocket will run standard Linux with all its comfortable bells and whistles, whether that is something like Pinephone or Librem, or whether I flip the table on the whole tiny-device-with-touchscreen compromise and just use something like a GPD Micro PC.

My 25+ years of Linux experience has shown me that it's better to compromise on polish rather than compromise on intended functionality and incentives. The Free ecosystem will gradually get better, and my familiarity with it will continually increase. The proprietary offering, which the Android ecosystem most certainly is regardless of the openwashing, will be forever limited by upstream's corrupt primary goal of bootstrapping subjects into Google's (et al's) commercial surveillance nightmare.

[p1mrx]

What is "4G deprecation"? You seem to be the only person talking about it:

https://www.google.com/search?q=%224g+deprecation%22

[mindslight]

The carriers are calling it the "3G shutdown" to downplay the fact they're obsoleting relatively new 4G LTE phones that simply don't do VoLTE. Hell, I even continued using mine without carrier voice service, but they eventually blacklisted the whole thing from the network.