said deloptes via tde-users:

| dep via tde-users wrote: | > I've never even seen a Linux cellular application. | | Besides the mainstream trying to do something for mobile devices, there | are few quite good projects.

Yes. I have a ThinkPad running Linux, and I want to make a phone call. So I repartition the drive to make a place for a Linux-based telephone operating system for amd64 and -- oh, wait, there isn't one -- after which I can make my phone call.

Rather, I meant a Linux *application* -- that's why I used the word -- that lets one make a call from the desktop, the way VOIP applications such as MySudo does, only employing the cellular network, enabled by cellular hardware in the device itself.

| For example MeeGo that later became Sailfish OS and was exported to | Russia as Aurora OS or the Android itself. | I've been using Sailfish OS approx. 10y now. It is quite usable and IMO | the best of all non Android OSs. They have also a small eco systems with | native applications.

Yes. I ran Sailfish on my dinky Planet Computers Gemini. I even got a Sony Xperia X to run Sailfish, which I licensed. (Which was slightly complicated in that they didn't sell to users in the U.S. so I had to install the Opera browser and use its built-in VPN to pretend to be in Europe so that I could send them their $50 or whatever it was.) Found that native applications were few and that the optional-at-extra-charge Android layer was kludgy and weird. Re-enablin the repositories after every update was a special annoyance. We were all pulling for it, but the Russian invovement was highly suspect, as was the heap of binary blobs. Finally settled on a Pixel running GrapheneOS, which is actually secure when intelligently approached.

I see that Jolla is still alive to some extent, trying to sell phones, computers, and even licenses for various devices, still not to the U.S.

But that wasn't what was talking about when I said I'd never encountered a cellular application for Linux. Which I still haven't.

said deloptes via tde-users: | dep via tde-users wrote: | > Yes. I have a ThinkPad running Linux, and I want to make a phone call. | > So I repartition the drive to make a place for a Linux-based telephone | > operating system for amd64 and -- oh, wait, there isn't one -- after | > which I can make my phone call. | | Now I understand. You want a program that can utilize the built in modem | and be able to make a phone call. | I have a Pinephone with KDE5 and it can do phone calls via the sim card, | but it is so heavy for the hardware and buggy in the handling that I | left it aside and do not know which software exactly does the work. You | may have a look there.

Yeah, straight-up Linux distributions for phones have always been kind of bad, and for some reason those who make phones for Linux seem bad at it, too.

(And I'm all for conservation, but if that's your goal, find something other than tin cans and plastic to put food in, because the telephonic burden on the environment is slim. If FairPhone would choose saving the world or building a great Linux phone, one or the other, they might do better. Pinephone at least has a good price for essentially the same phone, but it has a website suggesting that the company hasn't advanced beyond beta. Weird. Though my favorite is the Liberty phone, an entirely unremarkable low-end phone that runs Linux for the everyday low price of $2200!)

I flashed Ubuntu Touch on a Nexus tablet a few years ago and was totally underwhelmed. They listed a zillion applications, but they were all webapps and you can make those yourself in about two minutes: take an icon, list a browser with a website, and you're done. A couple of years ago Mint made big noise about an application that would do that, as if you couldn't do it already by following the instructions in the sentence above. It's like all the "Debian-based" distributions that take Debian, rearrange the hierarchy just enough to make them incompatible with actual Debian, cook up some unkillable little update announcer, and call themselves a distribution.

When I got the ThinkPad, the idea of wireless everywhere was a thing, as was ExpressCard (for those of us who didn't learn our lesson with PCMCIA). And lord, if I didn't fall for it! All of it.

Though there are apparently phones that run on Tizen, which has an Enlightenment-based frontend. I think that's where E gets most of its money, so it would be nice to see it actually working well, all of it.

But no, I'll stick with my GrapheneOS on Pixel and spend my money fixing up an ancient ThinkPad, and thereby keeping alive as much hardware as Fairphone sells in a month.

Thierry de Coulon via tde-users wrote:

I own both and I can tell you you can't compare.

The Fairphone works (with Android). The Pinephone does not "work", it's a toy. Does not keep a charge, no OS works correctly, many can't even... phone. The Pinephone seems mostly to want to be a hardware that let people develop an OS, but that OS - at least the last time I tried - is not yet even at beta level.

And the main issue is the eco system. Without eco system you do not have a market and without a market you are predestined to tail. Who is responsible for this .... well not only the government or the company that bribes through lobbying - a big responsibility lays on the developers and consumers too, but we can not blame the consumers that much. A consumer wants things that work and the linux developers communities in the past years are not quite able to deliver on this expectations for various reasons, so at the end I am also planning to move to Fairphone at the end.

Andrew Randrianasulu via tde-users wrote:

So, phone (cellular) hardware is patented, heavily restricted in other regulatory ways, and just plain complex (I am mot sure if there even full server-side 3G stack - research seemed to jump from 2g to 4g ... probably due to complexity in implementation we users never see ...)

3G is being phased out. They wanted to phase out also 2G, but found out that too many devices (from critical infrastructure) use the 2G network, so decision was taken to decommission 3G. 4G will follow decommissioning as soon as 5G starts working everywhere and ATM 6G is being developed.

And sadly bluetooth/wifi hardware and algos in similar situation, and even raw camera (not premade module talking via usb as usb video class device) and video de/encoding on hw can be challenging topics in themselves ...

IMO part of the problem is that big players bind manufacturers through OEM licenses and close the market, so that they have advantage. The microchips are getting so complex, that you can hardly reverse engineer them and you always risk to hit a problem when implementing. On top of this to have a profitable market (lets say for mobile devices) you have to have 300 million market, otherwise it is not worth producing anything except you sell it at the end for 2000+ ... so it is very difficult to keep up especially when everybody (big tech, governments etc.) is working against you.

So, "software is simpler", but only if you have enough engineers to throw at challenges ....

I had in mind to write a gui that helps you make phone calls through bluetooth. It uses OFono. I tried it through CLI DBus calls. I even started working on it, when I was not so busy few years ago. It works fine, but unfortunately there is no time to write GUI code :-/

Here is a citation:

In Linux, the Radio Interface Layer (RIL) acts as the interface between the operating system and the modem/radio hardware on mobile devices. It provides a standardized way for applications to interact with the cellular network and related hardware for tasks like making calls, sending SMS, and managing network connections. RIL implementations often involve a combination of kernel-level drivers (like USB serial drivers for modems) and user-space daemons (like rild) that handle communication with the modem

So this is the direction where one should go if utilization of the modem is needed.