> Doesn't that mean the core development is divided 14 times?

I don’t think it’s quite that simple. A lot of these competing mobile distros are using and working on the same libraries and porting desktop applications to mobile in similar ways. The biggest differences are usually the desktop distro that they’ve made their base and their desktop environment of choice. It’s common to see enhancements to an application or even core functionality like battery utilization or suspend improvements land in one distro and then be quickly adopted by others. As someone who’s mostly a bystander, the outliers seem to be the people working on the distros that aren’t ports of desktop Linux, like Sailfish, Nemo, and Ubuntu Touch. Of the main desktop-Linux-ports cluster the biggest divide seems to exist between launcher/DE development. Developers are split between the Gnome-based Phosh interface and the Plasma mobile interface. But a similar dichotomy exists on desktop. I’ll also throw in an honorable mention of sxmo and its wayland port, which I consider the equivalent to the underdog tiling window managers on the desktop; which is to say its only option good enough to consider using.

> I’ll also throw in an honorable mention of sxmo and its wayland port, which I consider the equivalent to the underdog tiling window managers on the desktop; which is to say its only option good enough to consider using.

It’s a funny thing indeed.

sxmo seems to be the only option not focusing on competing with Apple and Google on a “good enough for mainstream” ux.

I mean Phosh is nice and all. Having frameworks for contacts and calendars just like in Gnome is nice. Pulseaudio working just as on the desktop is great too. But the total experience still leaves the impression of a very subpar iOS/Android copy.

sxmo though, that has decided to not compete with Apple and Google on what they do best, but rather do their own thing and re-envision what a Linux smartphone should/could be.

And I like it. I like it a lot. It’s by far my favourite Linux smart-phone experience so far.

I just need a better phone to run it on, and the Pinephone Pro could be that phone.

New phone or not, I’d also appreciate if sxmo managed to rebase on/ship for Mobian too. The package selection for pmOS and Alpine is pretty weak in comparison.

I looked into sxmo and I think it falls clearly into the "clever hack" category and not in the "actually usable" category. There is only so far you can get on a touchscreen device without actually programming any of the apps to support touch. The lack of hardware acceleration in anything X related is also basically going to ensure that it always has poor performance, everything needs to be moved to use GLES based rendering.

> I looked into sxmo and I think it falls clearly into the "clever hack" category and not in the "actually usable" category.

Agree to disagree? Also I think you’re being somewhat disingenuous or uncharitable here.

sxmo has clearly had a unique vision for mobile Linux for power users and executed on just that.

All core phone functionality is available through regular, composable shell-scripts. And all major events can be hooked by simple user-controlled scripts in $HOME/.config/sxmo without any other alterations to the OS at large, no root required. It successfully employs a tiling window-manager by default to allow simple(!) mobile-oriented multi-tasking.

That’s quite something of its own, with no equivalent anywhere else in mobile space, Linux-based or not.

This is clearly a power-user enabling mobile Linux shell, and it’s making no excuses about it.

Sure there might be technical improvements which are possible at several levels in the stack. I’m not debating that.

That does however in no way take away from the vision behind it and how well that has been executed so far.

I honestly do not see what is so remarkable about it. It just never seemed to me like a unique vision but instead an effort to adapt some existing X11 tools to a mobile workflow. Which is a fine thing to do if you like those tools, but that's different from having some grand new vision.

To elaborate: The use of a tiling window manager with explicit workspaces doesn't really make sense to me on a phone since every app runs fullscreen anyway. The use of shell scripts doesn't really make sense since editing text on a phone is awful. I don't understand what the definition of "power user" means here either. What does this do that other phones and shells can't do? I can edit shell scripts in Termux on Android too, but it's still awful and unpleasant. Unfortunately I just wasn't able to figure out any reason to use it.

And just to be clear, I would not describe any Linux phone as a grand new vision. They're sadly all playing catch up. Maybe that will improve in the next few years.

Edit: I forgot to mention, the use of volume buttons to control a device with a touchscreen is pretty ridiculous. I mean, come on, you have that big nice touchscreen and you're not going to use it? Or has this improved recently where you don't have to do that anymore? Please let me know, thanks. Maybe I'll try it again if this is any better.

> The use of shell scripts doesn't really make sense since editing text on a phone is awful.

You don't have to write the scripts on the phone. You can ssh into it from your desktop. Or connect a keyboard and screen.

Yeah but that kind of defeats the purpose of having a phone, doesn't it? :)

I've seen some neat code editors on iOS/Android that do some fun things like:

- Have a special keyboard that contains only symbols for the target language

- Handles all editing at the AST level so you can't make typos

- Make heavy use of completions everywhere possible

There seems to be no reason we can't also have that here.

> Yeah but that kind of defeats the purpose of having a phone, doesn't it? :)

No, I would say it opens up infinite possibilities for useful customizations. Even if you aren't a programmer, someone likely writes a relevant script for your use case.

I mean if you have to transfer files from your computer or connect a keyboard just to do something on the phone. That's really inconvenient, I've tried to use my Pinephone that way and it's unpleasant.

The other platforms have figured this out, where the "someone likely writes a relevant script for your use case" becomes "someone likely writes a relevant app for your use case" and you just go to the app store and download it. No need to mess with scripts or plug in a keyboard, and you still get to have plenty of customization. I believe flatpak has started to add categories for mobile apps so that is bringing us closer to where things need to be.

Edit: I found the Purism blog post where they talk about this, the metadata should be usable by anyone doing packaging or building a mobile shell. https://puri.sm/posts/specify-form-factors-in-your-librem-5-...

I don't know, that just seems to me to defeat the purpose of having a touchscreen. I've tried to use those mobile VNC clients and the experience is really bad. It's something I would use if I really had to access some important files on my desktop, but I wouldn't try to log in to my desktop with a mobile VNC client every day just to do work.

good to know, appreciate the clarification! Glad to hear the work is cumulative.

>Doesn't that mean the core development is divided 14 times?

Honestly I don't think personally that it hurts things as much as one might expect. There is a lot of work from each distro that benefits all of the others. For example the Mobian dev was one of the first to really tackle the issue of battery life and their work benefited all the other teams.

That being said it might be better to have a few larger dev teams, but as long as their works contribute back to the community there is always benefit to the whole.

I'd second your assumption. Geary, MegaPixels (the camera app), Phosh (one of the few window managers), the modem firmware, and more are mostly shared between all the distros. Doubly so for upstreamed SoC firmware, battery efficiency improvements, and more that I'm sure I'm overlooking.

Frankly, it makes me question anyone who says desktop Linux is divided because we have options - at this point, it's more like there's no large corporate backer with a vested interest in making the desktop experience mainstream.

Asking for clarity since I'm one month into my Linux journey: Are you saying that most of the distros are really just more surface level changes rather than deep "divisions"? I've been struggling to understand.

They ship different things out of the box but can largely use the same exact software. They may update at different rates and thus have different versions of things. They are not entirely different operating system like Mac OS vs Windows.

Mostly. There are a half dozen different desktops to choose from. Same for ways to package software. Most of what a distribution does is make some choices and configure it. Most distributions learn from each other in some way. Most distributions fix things upstream so that other distributions can take advantage for their useful changes.

Most distributions are derivations and arise because some skilled people wished some core functionality would be different.

And other distributions started from scratch, but obviously don't want to do really everything on their own, so they took other software.

The result is that a lot of established software is shared.

SteamOS?

Good catch, thank you. While Steam currently prints money AFAIK, they're simultaneously 30 years behind Apple and Microsoft (as far as being the driving force of a desktop) and the reason Steam (and gaming on Linux) has come so far, so fast recently.