Preferences

I think this is the inevitable outcome of any movement of Linux to the mainstream (Purism has done something similar). As Martijn said in the article, PinePhone devices were operable with 25 different projects. That's 25 different variations of Linux fighting over market share. As Pine enters a growth phase for their business, the consequences of this are going to manifest as paralysis.

Improvements and advancements would stagnate due to the 25x duplicated effort, and resources would be lost in keeping those projects happy. Also, any potential user looking to switch would be deluged with options, which is what crippled desktop Linux.

While I do not understand enough about Pine to know why they specifically made the business decision to gut their dev community and go with Manjaro Linux, my guess would be something along the lines of Manjaro's widespread dominance as a top Linux distro backed by a powerful foundation. Pine is pivoting to what they have decided is their future: a full-stack hardware to software open source offering that in their eyes would have a better shot at cracking open the phone market.

They probably were aware of the consequences, but have bet on making it big and creating a new, streamlined ecosystem after extinguishing this one. It remains to be seen if they will succeed.


> 25x duplicated effort

People keep saying this, like too much effort is a bad thing. But OP makes it clear that there's very little duplication of effort in the Linux-on-phone/mobile/SBC community - most developments are shared community-wide. There are some technical divergences (Alpine/musl vs. glibc- and init-based systems for example) that can impact direct compatibility, but they're very minor.

Correct. This is non-sense speak by non-developers. Greedy non-developers.

they "sell" it as "25x duplicated effort" but in reality, there's 25x little tweaks to a build system, that give thousands of people zero effort to port their known platform right away. Now those thousands of people will have REAL effort to adapt their knowledge and existing ways to fit that one holy way enforced by the device true owners.

In reality it is "25x places where i will have to hide my plan for monetize this". Just like most other projects, greed always destroy everything the community help build in good will.

I would like to disagree with your characterization. I use Linux distros regularly at work (Amazon Linux), in school (Rocky Linux, currently studying), and at home (Ubuntu) when hacking together various projects. I've seen firsthand the issues that come with trying to get a distro to interop with Bluetooth, sound, and software not quite designed for it.

This is not a zero sum game: I believe we can have both an OSS approach to Linux while at the same time having a channel of commercial development that brings more adoption (and fun, hackable devices!). This "one holy way" and the multitude of community-based distros can coexist, in the same way that commercial software companies and OSS communities have already learned to.

> I've seen firsthand the issues that come with trying to get a distro to interop with Bluetooth, sound, and software not quite designed for it.

These issues crop up on bleeding-edge hardware due to different distros running differently-timed versions of the same underlying components. They fade away over time as software versions start supporting the formerly-bleeding-edge hardware across the board, and the issues invariably shift to the next bleeding-edge hardware release. This is not an immutable fact about the ecosystem, it's a consequence of wanting something to function before it's fully ready for serious use.

Right, this boils down to "Oh no! 25x many people are interested in our project. What can we do to reduce that to 1x people interested in our project?" If I can't easily boot my preferred distro, I don't buy your boards. That simple.
There's also a middle ground somewhere between 1 and 25.
I think that's a huge overstatement of how much duplicated effort is involved. The process is much more akin to:

* OS 1 finds a bug in Gnome, reports it and perhaps fixes it

* OS 2 benefits from pulling in the new code as well, fixing bugs

* OS 3 writes a driver for the camera and publishes it as part of their kernel

* OS 4 finds a bug in the camera driver they started using, publishes their fix

Yes, there's some overheard to running 25 projects. There's also a huge downfall to excluding 24 projects from contributing as first class members of the project. To boot, it's also a situation where the more contributions make the fixes contributed even more battle tested and beneficial.

tl;dr - OSS development styles don't map onto commercial development styles cleanly

I guess it comes down to that: will Pine64 take an OSS development approach or a commercial development approach? I've been swimming on the question of why Linux isn't more accessible to more people for a while, and have come to believe that a commercial approach is the only way Linux can achieve the work-out-of-the-box dream.

Commercial development allows you to afford to control the hardware, make deals with other companies, and pay people to build compatibility with your system (i.e. Nvidia), which is what Microsoft and Apple did to keep their position. Server distros like Debian, Ubuntu, and Redhat already have deep foundational and corporate backing, and are a joy to use.

There are definitely drawbacks such as vendor lock-in and all the issues that come with corporate vs community control of the software. However, I believe having a single center of development and revenue (to pay for the development), while at the same time having fully open source software and hardware is possible and would have a huge impact.

Historically, "works out of the box" has largely been a matter of low-level hardware bringup. That's one part of FLOSS development where there is already a natural "center of development and revenue", namely ODM's and OEM's. They just need to stop pushing hacked-together, barely-working downstream BSP's tied to a single software configuration, and start cooperating with projects at relevant levels of the stack. Pine64 is actually a lot better at doing this than your average hardware vendor, the OP is mostly complaining about relatively minor quibbles (though important quibbles nonetheless, because they directly impact OP's work).
Pine has been very much in the "we make hardware, community figures out software" camp, otherwise we wouldn't even have the discussions of "different things worked in different distros" etc.
As someone who works for Purism on the Librem 5, I have to say that I never had the impression of stagnation due to duplicated effort. Quite the opposite - we benefited a lot from work done on projects like postmarketOS, Mobian or Plasma Mobile, and I'm pretty sure others are benefiting from Purism's work just as well. A lot of stuff is interconnected and the push to mainline as much stuff as possible was certainly worth it long-term, as we can clearly see its fruits in our everyday work now.

This item has no comments currently.

Keyboard Shortcuts

Story Lists

j
Next story
k
Previous story
Shift+j
Last story
Shift+k
First story
o Enter
Go to story URL
c
Go to comments
u
Go to author

Navigation

Shift+t
Go to top stories
Shift+n
Go to new stories
Shift+b
Go to best stories
Shift+a
Go to Ask HN
Shift+s
Go to Show HN

Miscellaneous

?
Show this modal