Emancipating from GAFAM: challenges and solutions

[Dr_Cool]

Finished the most significant digital conversion of my lifetime (from someone doing digital since mainframes roamed the earth).

The main challenges that I faced were
(a) keeping compatibility with all necessary commercial and public service apps,
(b) preserving instant messaging functionality,
(ç) having UnifiedNlp location services, and
(d) minimizing short and long-term maintenance costs and time.

Given these four challenges, the effort centered around two axis:
(1) open system smartphones (chose Fairphone 2 with FP Open OS for the family) and
(2) an open source personal cloud (chose Nextcloud and DAVdroid).

In order to achieve (a) and (b) I tried to use microG for a while, but unfortunately it’s heavily dependent on few developers, and coding doesn’t keep pace with Google’s juggernaut.

Next step was to make a few concessions and move to Open Gapps pico. This showed to be a good compromise. I could freeze Google Play Store, a major irritant and invasive block in Google’s Android implementation, while reinforcing (a), (b) and (ç) and achieving (d).

The challenge then was to get out of some remaining GAFAM services, and the solution was to use a non-GAFAM cloud. Even though using an open source independent cloud provider is a decent solution, I preferred, as a user of a Synology personal server for many years, to build my own cloud server. I went with (2) above and it solved most of my problems. I have now my own cloud access to contacts, calendar (both Android stock apps), to do lists (Tasks) and notes (Synology’s own app).

Otherwise many adjustments at the margin, e.g., using Qwant and startpage instead of Google’s search engine, F-Droid Maps instead of Google Maps, etc.

The effort paid off nicely, I have now a robust Android-centered system without a trace of a Goggle account and compatible with most existing apps. The advantages:
(I) much improved data privacy and hygiene,
(II) much reduced battery consumption and better phone performance (setup isn’t dependent on apps that have to frequently phone home or exchange substantial data over the air to finance their business model),
(III) more control over features and suppression of irritant or invasive events,
(IV) better ownership of my own data and services and emancipation from whimsical decision making in places like Silicon Valley, and
(V) contributing to a more decentralized Internet and to many socially desirable outcomes such as less political, social and economic inequality, etc.

I believe that we have achieved enough critical mass in the open source movement to make this situation possible, but unfortunately the procedure is still too complex for the average user, this should be part of our next efforts.

It’s also unfortunate that I can’t do the same with my PCs because of my employer’s mandated dependency on Microsoft digital solutions. I’ll probably have to wait until I retire before I can setup an equally effective solution at the PC level, but given the more intrusive nature of smartphones and the ability to use Mozilla products on a PC I rate my own digital hygiene currently at an 8/10 compared to a 2/10 just a year ago.

[Roboe]

@Dr_Cool

(I) much improved data privacy and hygiene,
(II) much reduced battery consumption and better phone performance (setup isn’t dependent on apps that have to frequently phone home or exchange substantial data over the air to finance their business model),
(III) more control over features and suppression of irritant or invasive events,
(IV) better ownership of my own data and services and emancipation from whimsical decision making in places like Silicon Valley, and
(V) contributing to a more decentralized Internet and to many socially desirable outcomes such as less political, social and economic inequality, etc.

(I and II) Even if you don’t see icons for common Google apps on your phone, you are still using Google Mobile Services through third-party* apps. So Google is constantly calling home, specially for their centralized push notification service —that’s how it works.
(III) If a third-party* app load a map (for example), you have little to no control over how it’s loaded. In particular, third-party* apps developers are encouraged by Gobble to use their proprietary Google Maps embedded maps.
(IV) Improved —yes, by using own hosted solutions for some things— but not emancipated.
(V) Great.

*= except third-party apps on F-Droid. They follow an strict inclusion criteria that specifically forbids this kind of “leaks” caused by proprietary services and libraries. Similarly, few developers outside F-Droid (in a case-by-case basis) may follow similar criteria for their apps.

IMO, title may be missleading —you’re still feeding and being dependent of Gobble. Of course, less than before, that’s indiscutible. You chose a conservative approach —good for some use cases, like people who directly depend on proprietary apps to do their work properly—, but it should be seen as a first step in breaking free from Gobble.

I applaud you for doing this and taking the time to write it down, though, and I also encourage you and other readers to keep following new developments and further improving their devices with the time,

[Dr_Cool]

Thanks for the comments. Indeed, what you say is true, in particular the dependency on FCM or GCM, currently there isn’t any reasonable alternative available. I tried for example Telegram in their proprietary instant messaging mode but it results in substantial and quite evident increase in processing and battery consumption while the phone is sleeping. Apple and Google maintain a tight hold on the World Wide Web because of being infrastructure monopolists, yes, it isn’t even a duopoly since users can’t easily move from one service to another. A problem for which solutions will need to come from beyond the open source community, it will need for example European governments to establish framework parameters at the EU level.

It’s true also that many app developers are totally dependent on Google services, reason why I finally had to use Open Gapps. But this solution significantly hurts Google’s business model because it renders Google much less able to use its accounts as hubs for user data extraction. For example, the app may download a map tile from Google, but Google is unable to associate this download event with a Google account that should be phoning home from the same IP. Even worse for Google if the app uses UnifiedNlp instead of Google location services (this is true for most of the apps that I use). It’s in reality a gigantic gap in Google’s model, and I expect that, when things get tough for Google, it will try to close it, but if it didn’t do it by now it’s because this is very hard, maybe not even feasible for legacy reasons. These monopolies fortunately can’t do all they want, they have to compete for example with their own “nicer” versions that existed in the past.

Emancipation in this case is an objective, evidently I’m not there. As you mentioned, one of the ways to further it is to prioritize F-Droid apps. This is why I concluded by saying that, according to my own estimates, I went from a 2 to an 8, but not to a 10, so I think the title is appropriate.