I also use Linux with my bank. All banks I know of in my country require you to install an invasive "security software" (warsaw), even when you're using the web browser; some also require the phone app to be used as a second factor (for instance, one bank generates a dynamically changing QR code after you enter your password; you have to open its phone app and scan that QR code, which the phone app uses to generate a number to be typed below the QR code).
For the government, it's easier, you just have to install the ICP-Brasil root certificates (since they haven't been accepted yet by Mozilla).
It's not just that Linux isn't going to break the app store duopoly in your case, nobody is going to, because your government is effectively reinforcing that behavior. Forcing people to install root certificates on their devices is a problem on a whole 'nother level over just commercial incentives.
I just don't see how any technology or market-based solution will solve that problem. What happens when the government starts deciding that its authentication software won't work on rooted devices? What happens when the bank decides they need contact/GPS/camera permissions on your phone for "security"? I mean, that's not an app ecosystem problem at that point, it's just full-fledged required malware from your bank.
In the US (for all of the many, many faults with our trash-dump of a banking/federal ecosystem) at most I've only ever been asked to do 2FA using completely open standards, and the worst problem I've run into is that my bank's website won't load over some VPNs. I can even disable analytics on my bank's website and it'll still work.
But getting support on a Linux phone OS for what sounds to me at first blush like malware and spyware isn't really the best solution to what you're describing. The problem isn't that the malware isn't available on Linux, it's that you're forced to install malware on any of your computers, including iOS/Android.
The government isn't forcing anyone to install the ICP-Brasil root certificates; if you don't, the government websites will just show as untrusted in the browser (since they're signed by one of the CAs which chain to the ICP-Brasil root), but other than that they should work the same. Things might be different if you are authenticating your identity to the government with a certificate stored on a smartcard or USB key, since these certificates are also signed by a CA which chains to the ICP-Brasil root; I don't know if adding the ICP-Brasil root to the browser is necessary to make that work.
As for the banks, AFAIK the government doesn't mandate how they approach their online banking security; their use of that invasive "security software" is probably mostly path dependence at this point. It was very common to see computers infected with malware designed to steal online banking credentials from the browser, and that "security software" was supposed to block that kind of malware (for instance, by blocking the browser accessibility APIs whenever the user goes to an online banking site). Things are already better than in the past (that "security software" used to be Windows-only, for instance, so Linux users couldn't use online banking).
And yeah, all the bank apps I have on my phone require contact/location/camera permissions (the camera permission is unavoidable since they have to be able to read the QR code for PIX payments, and boleto bancário barcodes), and for at least one of these apps, I strongly suspect it's using the location permission for security purposes (the location indicator always lights up whenever I log into that app). It wouldn't surprise me at all if it also rejected rooted devices.
My issue isn't that those apps aren't available for Linux. Even if they were available for Linux I wouldn't install a bank app with those permissions. I wouldn't install that on my primary Android phone (I guess there's a good likelyhood I couldn't, since I run rooted Android).
I'm not sure what I would do in that position. Maybe get a separate phone just for banking? I don't know, it's really tough. I sympathize.
But yeah, the Pinephone is not going to solve that problem, banks would need to move over to open standards for authentication through apps that are trustworthy. Just releasing their apps for a smartphone wouldn't be enough. At least any "security software" that can run on desktop Linux can also be run on your phone, assuming it doesn't freak out over the hardware, and assuming you're willing to tolerate it.
Sure, if you want an app for an open source platform like Linux, that may take a lot longer. I think there is already a common platform for this, but not many banks offer it.
Is this meant to be taken literally? I would be surprised if e.g. mBank didn't work with SMS as the sole enabled 2FA.
Personally, I would rather have an extra Android phone for cases like this, as well as for the times when one has to make an actual phone call, though. I wouldn't expect the call quality on the pinephone to be anything to write home about, even if the software was all mature already.
(I use Linux exclusively on the desktop so I know what you are talking about, but it just isn't going to work on a phone in the same way).
Currently, I don't use a banking app because I don't trust my bank to have that level of access to my phone with a proprietary app, and because my bank's app isn't particularly good in the first place. So I'm not really losing out by using Pinephone. In fact, I might have more functionality after switching, since Android development is a mess and on a Linux platform I might be tempted to build something quickly for myself that periodically pulls down the CSV data from my bank and does a few things with it.
I'm not saying that proprietary devs are going to start mass-building apps for this, I'm just saying that for a segment of the population, this may turn out not to be the problem you think it is.
----
Apps on my phone that I need replacements for:
- NewPipe
- Podcast app/music app
- Maps
- AndOTP
- Signal (and/or Matrix)
- Orgzly
- Good browser support
- Maybe a few misc utilities, probably something I'm forgetting
Do I have confidence that those apps or equivalents will be ported? Honestly, yes, I do. It would not take an insurmountable amount of work to build a Linux ecosystem that was more functional than my current phone setup.
You can find a bigger list of apps in https://www.plasma-mobile.org/. There is a few still missing that I will try to add later.
- NewPipe > FreeTube, AudioTube
- Podcast app/music app > GNOME Podcasts, GPodder/Lollypop, Elisa
- Maps > PureMaps, GNOME Maps
- AndOTP > Keysmith, GNOME Authenticator, Two-Factor-Keychain, Numberstation
- Signal (and/or Matrix) > Axolotl / various, Nheko feels most complete currently
- Orgzly > Emacs maybe?
- Email > Geary (I still use Evolution though)
- Good browser support > Angelfish, desktop Firefox has been adapted nicely, once Chromiums Wayland supports matures it will also be a viable option (it’s very fast, but has annoying issues with text input currently)
To develop apps for Android I have to set up an emulator, dev environment, learn a bunch of extra ropes, package everything. But I can throw together some bash scripts in a day, it really opens up the possibilities to solve a lot of my own problems.
Compare that with something like Windows Phone. Not having apps there is a serious problem, because I don't necessarily want to pull out C# and build an entire app to fix every issue I have.
Can also be done with Android app via Anbox: https://social.librem.one/@dos/105119691527269967
That's part of the point of using Linux -- not creating an ecosystem from scratch.
Virtually all of my computing outside of my phone is already done on Linux devices, and the setup works for day-to-day computing. Support is good enough for me. What makes you think Pinephone going to have worse support than my current Linux desktop?
My hope with Pinephone is that a decent influx of Linux phone users will improve the quality of desktop software on Linux as well, since it forces devs to stop ignoring touch screens and to start thinking about responsive design. Ideally, the desktop versions of many apps should scale on interfaces like this without needing to be rewritten. But regardless, I'm still not particularly worried that my bank website is going to stop working.