Hacker News Reader

Comment by djent | Hacker News Reader

It is frustrating to me that despite the blunt verbiage about it being a beta device for software and hardware developers, people even here are griping about it being not ready for everyday use. If a free and open-source mobile operating system is missing software support for a feature you want, please contribute to its development.

View on HN

How many hundreds or thousands of enthusiasts are needed before it's suitable to be a daily driver? I see the fragmented approach on the software side as a major hindrance to the goal of achieving a device that can be used by a normal person as a reasonable replacement for other smart phones.

I check in on it every l every 6 months or so, update my pinephone and give it a few days. So far, it doesn't meet reliability or usability for serious use.

It's a high caliber foot gun of a methodology - if they corralled all of the effort and enthusiasm currently scattered amongst a dozen (!) different os's and communities, they could establish a high quality functional baseline.

The hardware is great. The mission is great. The software is fragmented and scattered and missing the force multipliers that focused collaborative development could bring. There's not enough to riff on, toomuch goes wrong in complicated ways, and so the device is an exercise in frustration instead of a passion project.

> How many hundreds or thousands of enthusiasts are needed before it's suitable to be a daily driver?

Zero. What is needed is a few hundred developers writing actually code to make it useful. The enthusiasts are at best nice to haves. You can blur the line if you write good bug reports, but right now the pinephone doesn't need end users.

The above will change over time. Millions of users would be enough to force carriers to support activating it on their network, something that developers cannot do (at least not legally). Likewise large numbers of users would be enough to those banking apps written. However today millions of users wouldn't make any difference to the important parts of making it work (except in so far as users sometimes become developers)

Enthusiasts with wallets are another story.

All my banks function just nice on my phone without any apps. There is one magic app that makes it happen, it's called a WWW browser.

> How many hundreds or thousands of enthusiasts are needed before it's suitable to be a daily driver?

I know your question was rhetorical, but when I think about this, it took close to 20 years of Linux maturing on the desktop before I felt comfortable handing a family member a laptop running Linux. As an enthusiast, I started running it much earlier, but everyday tasks like "printing" or "using Wi-Fi" prevented it from being a daily driver.

Personal computing, though, had a much deeper hacker culture, that encouraged pushing the hardware to its limits, and doing new and novel things with it. Progress stemmed from that (Windows including BSD sockets, for instance). I've never seen anything similar in the phone ecosystem - as soon as you play outside of the walled garden, you drop off into fringe territory. And the PinePhone, as much as I love it, is by definition outside of the walled garden.

I'm not sure, given the culture around phones, we might ever get to the point of it being viable.

I don't know if fragmented is the right word. Just like desktop Linux, mobile Linux has pretty much settled on one major standard (Phosh a la Gnome) and slightly less popular alternative (Plasma Mobile a la KDE). Sure, you can go about customizing it to your whims, but fragmentation isn't nearly that bad, especially since the distribution options are so much lesser.

The problem really is getting the commercial software people want on the phone. But if you just want to use it as a phone (e.g. no special apps or mobile games), it does most of that in a pretty streamlined manner (albeit, rougher than the commercial options).

TBH I think having a flutter port for this operating mobile system is probably the most likely way for it to succeed. That way you can do a flutter app and it will work on this too, while still having mainline OS support.

Then you would need to fill in some common services like push and such.

Or run an Android or iOS emulator. (Note: like Wine runs Windows software on Linux in a fully legal way).

I think that is actually counter productive, since the emulated apps will always be shittier than their native counterparts, and you need to constantly deal with telemetry shims for Apple & Google lest the apps wont work properly.

It's desktop Linux, so you can do this now! I know of at least one Flutter app available as a Flatpak on ARM Linux (FluffyChat).

Yeah it's really close already and has adoption in the mobile world, which is why I would suggest the community adopt that vs something else like mobile gtk. I dunno how optimized flutter desktop for linux is for the mobile version, maybe all it will take is a view port change and implementing some more mobile orientated apis like location & orientation changes.

The most likely way to succeed would be to jump on the PWA bandwagon, IMO.

I think both would be necessary.

One problem is even governments support only the duopoly of smartphone makers. For example, try downloading a "Covid App" for a Linux phone.

https://en.wikipedia.org/wiki/COVID-19_apps

In my state, the "app" is actually a website (https://myirmobile.com/).

I'd be quite happy to _not_ be able to download "Covid App", whatever that is, on my phone. So, what's a bug for some is a valuable feature to others.

It's the slippery slope that's the problem.

It's kinda weird anyway, for showing a QR code it being an app adds no benefit. You can just print the QR code and keep it in your pocket. It's valid for 1 year in most places.

On iOS PassBook would be the ideal place for this, not a separate app.

I do have mine on my phone but just in case.. Here we don't need the code unless we go to the disco, and I'm too old for that anyway.

Either way I'm vaccinated so code or not I've done the right thing :)

Even though I'm vaccinated, I have no interest whatsoever of proving it with any kind of "code", "star of David" or whatever the fuck Faucists want me to prove it with.

It’s less bleak than one might think:

For tracing, at least Sailfish OS has an app called Contrac [1]. I looked into porting it over to other PinePhone OS, but with the CRUST and deep sleep, it would not really be useful there anyway (as it would not be active most of the time).

Regarding QR certificates, there’s Vakzination [2].

[1] https://www.flypig.co.uk/contrac

[2] https://invent.kde.org/plasma-mobile/vakzination

> How many hundreds or thousands

Hundreds of millions. That's how many you need for mainstream apps to be built.

Even Microsoft couldn't get past that hurdle and money is not a problem for them.

Android compatibility might help somewhat but it will undermine the usefulness of the open platform by introducing the very things it fights against. The OS is not the only problem. Apps can still datamine the Android layer then. You won't be much better off than with a privacy focused Android fork.

Android is a much better base to start from anyway. At least that has a somewhat coherent security model.

> How many hundreds or thousands of enthusiasts are needed before it's suitable to be a daily driver?

I just need a WhatsApp client. Enthusiasts once made one but WhatsApp started banning their users.

> I check in on it every l every 6 months or so, update my pinephone and give it a few days.

Why did you buy the phone? They pretty much shout that they are only targeting developers who can bring about your dream. They never promised what you're expecting of them.

> if they corralled all of the effort and enthusiasm currently scattered amongst a dozen (!) different OSes and communities, they could establish a high quality functional baseline.

If Apple and Microsoft combined efforts on Mac OS/Windows, they'd do amazing things too - but that's not going to happen either: once formed, organizations expend significant energy in self-perpertuating. Perhaps you could pick a promising project (or a couple of complementary ones), and support them technically or financially.

As a owner of many pine64 products I wish there was a clear path to learning how to contribute. My last time coding was C++ class over 16 years ago and I wish I had some sort of idea on how to get started with hobby programming in my spare time as an adult.

Edit// Especially since I've jumped on the ARM bandwagon and replaced my desktop with an M1 mac and my laptop with a Pinebook Pro.

Heh. Contributing to open source is....interesting. Pine64 really seems to only do the hardware aspect, everything else is up to others.

I guess I would ask: how do you want to contribute? Is there a specific niche you want to help fix? That's how I got involved, I wanted MMS (and later visual voicemail) for my Pinephone.. Or perhaps there's a specific project you like that needs help, or theres a particular bug that annoys you in a FOSS project. Those may be the best ways to get involved to help.

Mostly it is you just need to dig in and make false starts until you figure it out.

It certainly depends on what feature or what app you would like to work on. Mobian and Manjaro are both on GitLab, and the public is able to create an account. Issues there may be marked Help Wanted or similar.

In regard to motivation to continue programming, that can be the most difficult. Overall if you are proud of your work and find it useful yourself, you will probably continue the habit of contributing and supporting your code.

go smaller, if u use some open source software check it's repositories and community forums and find something that people don't like to do

documentation is definitely a candidate also automated tests

btw, being on arm makes Pinephone development easier, since you don't have to cross compile.

Just pick something that bothers you or you think is missing and start working on it. I'd recommend something application, as opposed to system, level to get started. That way you can develop using something like Python and not worry about lots of other dependencies on whatever you're working on.

There generally isn't much in the way of formal organization in the open source world... just scratch an itch that you have. There are relatively few people working on things specific to Linux on mobile so even a modest contribution can have a big impact. For example, the developer of the Megapixels camera app, which is a vast improvement over what existed previously but still needs much work, just wanted a better camera app.

Create apps! This is the most needed. We created a tutorial for Kirigami here: https://develop.kde.org/docs/kirigami/. There are also a few tutorials for GTK/libhandy with e.g. https://tuxphones.com/tutorial-developing-responsive-linux-s...

Also join the developer channels on matrix: #plasmamobile:kde.org (not sure that the address for the GNOME one)

Cool! This is the kind of thing I've wondered about for a while. This looks like some handy info.

Is there an option to use a higher level language like Python and these QT libraries still? I don't have a ton of time to contribute, but this might be a way forward for me someday.

EDIT: Your 2nd link is literally what I was asking about, although it's using Gtk3. I'm not familiar with either in depth, so I'm not picky there. That's really useful. Thanks for this!

I feel like programmers have become "soft". I bet the old guard is okay with something like this - a lot of C, C++, assembly skills - this is what it meant to mess around with computers!

Now, you can just be modifying CSS and call yourself a programmer. You get one of these devices, and you're sorely disappointed!

As the complexity of our every day devices increase, less and less people are capable of contributing anything meaningful to an OS like Linux, let alone in their free time. It also doesn't help that the abundance of high level languages discourages learning about computer internals in younger generations.

Or that we indeed became softies lol.

Those harder classes used to be referred to as weed out courses. Incoming class of 100+ students to Assembler. After 2 weeks, 60%+ drop the class.

I'm currently TA:ing my uni's Introductory Computer Engineering class (from logic gates to assembly). It's not as much of a weed out class as you'd think and yes, it's the first class they take, along side some introductory python. Most of them do seem to hate the assembly part though personally I found the part where you have to actually make an instruction to be much harder when I had to take this class

Language is very small part of it. For example me contributing a anx7688 driver for pinephone to make convergence work was C coding, yes, but also reading through type-c spec, battery charging 1.2 spec, usb-pd specs, alt-dp specs, and figuring out how it all works from 0 knowledge, to tie all that together on a quirky HW design, with several hardware bugs that I had to discover first, and non-cooperating PMIC/and type-c controller, on 3 different pinephone HW variants.

C coding is the easiest thing. Hard part is figuring out what needs to be done and getting quite detailed understanding of how everything works on HW level, lots of trying and testing with various USB devices in various scenarios. There's also a lot of reverse engineering, because no HW vendor cooperates with random fucks from the internet and gives them free support. :)

So drivers are definitely awesome. But at some point adding some usability for “killer apps” on the mainline phones might pull in significant users.

Killer apps for me are physical keyboard, GUI bootloader that can do multi-boot (so that I can switch between one of the mobile distros and a Xorg based i3wm desktop, because there's no way mobile distros will ever have reasonable performance for convergence use), convergence with accelerated video playback and a reasonably smooth web browsing, possibility to run Arch Linux ARM as is with access to the entire package repository of software (that is not ad laden or spyware by default) on my fingertips, and full control over the security of my phone by being sure I can run my OS without the CPU ever touching any code on any modifiable storage inside the phone,... :) That's already there with original Pinephone.

The pool of programmers has grown. There are more good programmers than ever, but the barrier to entry is also lower. It isn't necessary to pass through that kind of trial by fire anymore.

Yeah, the barrier is lower, many people can write Electron apps now.

I don't think the elitism here is necessary or even the root of the issue of code contributions. The apps currently present in these PinePhone operating systems were developed and are supported by knowledgeable and hardworking PinePhone users, but the fewer developers there are, the more limited the support is.

> a lot of C, C++, assembly skills - this is what it meant to mess around with computers!

I can't believe this is getting upvoted on HN. What kind of gatekeeping elitist bullshit is this?

Though a bit elitist, his comment isn't discouraging anybody, and does hold some truth in that getting closer to the hardware allows you to do things you can't do otherwise.

I feel like it's unfair to disparage him as harshly as that.

I don't see anything elitist. Anyone can learn assembly language, C, and even C++ and Rust with a bit more dedication.

Anyway, much of the work getting these phones ready as daily drivers is in getting apps that run on them mature. There are lots of languages adequate for apps.

Or learn Swift or Kotlin, and target iOS or Android instead.

Or even, you know, not.

What's elitist in learning a craft properly? Working with a higher level language doesn't mean that learning C, C++, algorithms, data structures and a bit of math isn't terrible useful.

Part of the problem is that the meaning of “beta” has gotten extremely diluted, and is often used as a CYA rather than an actual indicator of where the software is in the dev cycle. Gmail was in “public beta” for what, five years? Game companies routinely give out “alpha” access that’s really more of a pre-release test and hype tool.

I bought a "beta edition" regular PinePhone and it seemed anything but "beta." Random crashes after boot every 30 seconds, even with "workarounds" such as patching the memory frequency (really?). This may have been a hardware defect but one of the nice things about calling everything a "beta" as an excuse for the flakiness I'll never know. It was essentially a paperweight, not something that would enable me to become a "contributor."

Sounds like a beta to me. Perhaps you're thinking of a release candidate?

From my experience a “beta” should be ready for daily use but some bugs should be expected. If it’s really unstable call it ”alpha”.

The product page doesn’t include the word “beta” anywhere.

I wouldn’t even call it a phone if it can’t act as a phone out of the box, let alone use the “pro” moniker.

Why is it not labeled as a development board with a touchscreen and cell modem?

It says it's available for developers.

Unfortunately I'm fully booked, and would gladly pay an extra hundred or two for someone else to do it. Not clear on who that is yet.

Unless you are talking about hundreds of thousands of dollars though your money isn't really useful. Developers are on the higher end of the income curve and so few people alone can afford to hire them like that. You can donate a few hundred so several projects, it will make a difference, but thousands of other people need to do the same before they can hire even one person (and the first people hired will do more IT or administrative work that needs to be done)

It's done by purchasing the device, many folks purchasing it, which I already did. In fact someone might consider starting a company (haha), or funneling the money to a single engineering source.

I happen to work for a company that employs a team of people to do just that, which allows me and others to work on mobile GNU/Linux ecosystem full-time. It's called Purism, and the product that pays for it is called Librem 5.

You mean the thing called PRO isn't ready for professional use?

Reply on news.ycombinator.com