One of the main reasons I ditched the CAT B35, and KaiOS more generally, was the fairly woeful implementation of predictive text. On a phone with a keypad the main mode of text entry – for SMS, but also in the browser and more widely in the phone operating system – will be some form of T9 text input.
It’s fairly mind-boggling to me that with 20 years of advancement and thousands of times more computing power, modern phones struggle to get as efficient key-based input as 90’s Nokias.
The purpose of a modern phone is of course communication: not only via the cell radio, but over the internet.
Naturally, this is a vital tool for the user. But the predominance of always-on network connectivity has lead to increasing abuse by providers of phone software: both Google and other creators of the operating system, and third-party app creators. As far as they are concerned, the phone’s true purpose is to (i) gather information about their users and (ii) serve advertising back at them based on this information.
Both these functions – antithetical to the users needs and wishes – rely on the internet. The baleful combination became possible with the rise of widespread mobile internet. In this post we’ll take control back over our network functions, so only the components we says can ever send or receive data on our phone. After all, we pay for it!
Don’t Be Evil is now long forgotten. It’s almost a truism that the purpose of Google is to harvest our data so we can be advertised to. But are we stuck with Google on Android?
We all know that Android is a Google project – but things are a bit more complicated than that. Google does not in fact control Android; rather it is an open source, and freely licensed, project worked on by a consortium (and is merely sponsored by Google). In fact, the majority of Android and the applications it runs is built from open source code and tools.
Unfortunately, Google have in practice made it as difficult as they possibly can to use Android without involving the Google ecosystem.
Google have attempted to insert themselves into almost every process in your phone, hoovering up as much data as they can, and claiming everything will break if they are removed.
This isn’t true. Let’s de-Google Android.
Six ways in which Google make it as difficult as possible to avoid using their products in on your Android mobile device:
As I described in the backstory, hardware-wise the Xiaomi Qin F21 Pro was the perfect form factor for me. The power of a smartphone, but actually designed to be a focused communication device, rather than a hamstrung computer.
There are thoughtful debates about the security pros and cons, but the prerequisite for this level of control is rooting the phone.
However, on the software side it needed a bit of work. The phone arrived with little bloat and mercifully no Google components – but there was still a fair bit of Xioami stock software, including the notorious browser which reportedly was spying on Chinese users for every page browse [I have no idea if this is the case in the internal Qin].
I’m a later adopter to smart phones. My early experiences confirmed just about every bad thing I’d heard about them. I was astonished, for devices of such putative hardware power, just how ridiculously and deliberately limited they were in software:
A very simple and ugly design principal underlies standard Android phones: effectiveness of advertising penetration means everything, the wellbeing of the users means nothing. The users are only a commodity input to the only reason Google exists: advertising revenue.
For a long time now I’ve been fairly comfortable simply stating
I plain don’t like smartphones
I’ve gone into this in more detail elsewhere – but the short version is:
However, I need a telephone of some sort. And while I try my utmost to eliminate it, there is some software which (grudgingly) I’m compelled to run outside of a PC.
A series of blog posts going into great detail on how I left behind large-touchscreen smartphones and Google – and set up a lightweight, reliable but powerful Android phone that works for me rather than vice versa.
Some useful resources that inspired this work:
Most operating systems have some sort of branded animation or placeholder while they are loading. In Linux, this niche is filled by plymouth. Unfortunately – and this is far from an exclusive scoop – plymouth is a pain in the arse and doesn’t work very well. I’ll never get back the day of my life I spent getting this working – but maybe I can save someone else the time (or at least the bother).
What follows are my attempts to get some sort of loading screen working for Gentoo on my MacBook Air before X/LXDE loads. This was a ludicrous amount of effort to get something so trivial working.
Unless you have a lot of time to kill, I wouldn’t bother with plymouth at all. But for the curious, here’s what I did.
The webcam built into the MacBook Air is quite odd in a number of ways. It appears to use an unusual hardware configuration involving a wifi-disabled Broadcom BCM15700A2 chip – and it was the one piece of hardware that didn’t work out-the-box on the Linux Mint installation. I also identified it as the culprit that was preventing deep CPU package sleep states (and hence full battery life) under Gentoo.
It isn’t too difficult to get it working properly – but it does require a custom firemware driver and activation of the right kernel modules.