So App Store is a monopoly on Apple devices, and on Android based devices there's plenty of choice ranging from bad for privacy (Play store) to fully FOSS-only (F-droid).
On iOS, if you don't publish the app on App Store, 0% of your users will be able to install it, because there is simply no alternative.
For us (well, atleast me), getting an apk from github is not a problem, also not clicking three times to update the app... "normal users" usually don't do this, and some even get scared at the scary messages on phones, warning users about "alternative sources" and "play protect" and other stuff.
I'm not really arguing that this is a great solution for everyone (or that sideloading should magically mean that Android isn't being noncompetitive), just pointing out that even insufficient solutions can really make a really big difference for some people.
I want the sideloading and updating experience on Android to be better, and there are some very real frustrations that come from that -- including that there are great apps like NewPipe that I can't really recommend to everyone I know because of the complexity of installing them. But in contrast, I don't think I could even use iPhone myself because these apps are so essential to me, so I'm grateful that at least I have a phone where it's possible for me to jump through these hoops.
To your point though and in regards to the person you were replying to, you're right that none of that changes anything about whether Play Store has what is essentially a monopoly over other Android storefronts. I would love to live in a world where I could recommend NewPipe to people without feeling like I also need to borrow their phone and sit down with them and help them install it.
----
> For us (well, atleast me), getting an apk from github is not a problem
Pro tip which you might already be aware of, but F-Droid supports custom repositories, and the NewPipe team maintains their own repository that gets immediate updates. Installing NewPipe from F-Droid directly isn't something I would recommend because the app needs to be updated so frequently and F-Droid is slow to pull in those changes. But if you use their custom upstream repo, you can still have F-Droid manage the updates/installation while getting basically direct updates from the Github repo.
This would of course be better if Android allowed the F-Droid store to auto-update apps, but...
I vaguely remember hearing that at some point Android was going to allow this, but then I haven't heard anything since and my version of F-Droid still requires manually clicking an install button, so I'm not sure if Google ever actually followed through.
Did some more digging and it looks like F-Droid just hasn't been updated yet to take advantage of it for eligible apps (https://gitlab.com/fdroid/fdroidclient/-/issues/2316).
99% of app users get an app by clicking on a link that was sent/recommended to them.
So more accurate to say "100% of users will never find your app on a store if it's not directly on a store."
I work on health/social service apps for state and local municipalities. The conversation about putting them on the stores ends when we explain that the stores don't allow such apps. Which is a blessing in disguise ans PWAs are far superior in most every way.
Most people just look at the "trending" stuff in the main playstore window, or search the store with general keywords (eg "menstrual calendar"), and try the first few hits.
Apple simply banned them. That’s why I cannot equivocate Apple and Google’s market dominance, Apple is just so much more obscene and over the top, Google at least has some more shame.
Epic sued Google too. They're going to trial in January 2023.
But then they weaponize inconvenience. You can run the app, but if you want Google Play, they stick you with their terms. SafetyNet blah blah. If you don't, the bank owns the copyright on the bank app so third party devs can't give ordinary users a convenient legal way to install it.
So third party alternatives get friction on purpose and can't gain users. And without users people don't develop native apps.
Compare what it would take to get iOS apps to run on a Linux phone.
Blackberry, Jolla and several other Linux + Android attempts have proven that being easy isn't enough.
You can, of course, access your bank from a browser.
For now. Do you think the W3C would resist if banks tried to standardise a way for sites to determine if you are running with Secure Boot enabled? This would be part of a system to prevent browsers from spoofing their user agent string / version number.
They ignored the protests of those opposed to DRM, and I can't imagine there being as much objection to banks trying to protect users from browser vulnerabilities and kernel-level keyloggers.
Of course some Free Software advocates would again "cry wolf", predicting that banks would start to block requests from "unapproved" OSes, but those claims would be dismissed as scaremongering, right up until it started happening (and the wolf started eating their sheep).
Linux on phones isn't even at "good enough," let alone compelling.
Its interesting because you can really see how we end up with dictators in politics. Doesn't need to be a coup, people voluntarily give it up for small conveniences.
Alternatively, it's not that important and as such the market being a "fiefdom" is irrelevant. In which case suggesting people switch to alternatives is also irrelevant because it simply doesn't matter.
I'm not sure what you're trying to say here. Even if you think that the status quo is problematic, I don't think that suggesting that consumers switch to something that is terrible is a pragmatic solution.
And yeah, Meta sucks.
Also, competition law seems to be way too simple. We should have at least 10 similar-sized competitors in this field, yet we are stuck with this duopoly.