De-googling my Life

Preface

This is a one-off piece not intended to be the second article referred to in my home media server upgrade article, but rather a collection of my thoughts and process surrounding the circumstances that led me to “De-Google” my life. A bit of a disclaimer, I made all of my decisions considering what would be best for my situation: This is not meant to be a guide. While it can be used as a framework, copy and pasting things verbatim will almost certainly not work out how you would expect 😉. With all of that out of the way, let’s begin the story!

Introduction

I started my journey to disconnect myself from Google as much as possible around 1 year ago at the time of writing this piece. Looking at where I am today, I’m fairly content with the progress I’ve made so far. Before I started this project, I used Google for everything. Search, entertainment (in the form of YouTube), an email provider, storage solution (for random files and my photo gallery), Calendar provider, etc. You name it, and it was very probable I was using a Google service to fulfill the need. Now, don’t get me wrong, Google is very convenient and is perfectly adequate for 99% of the population. It’s just for people who tend to overthink about their privacy…. it can be an absolute nightmare. Putting all of your eggs into one basket is kind of scary to me. For example, what if Google goes down? This is a very, very, unlikely scenario, but it does happen. About once or twice every few years it seems (it doesn’t seem to be published information, so take this with a grain of salt, it’s my best estimate). No computer system, no matter how redundant, is infallible. Things will happen. Last time Google went down was the end of 2020 for about 1 hour. During this time, it’s safe to say a good portion of at the very least the U.S. had to stop functioning. Imagine: business emails powered by Google, not able to send or even look at new messages. Even if they could, it wouldn’t matter, anyone using Gmail had this same issue. Almost no one could get to their email, nothing could be done. Any data on Google Drive is suddenly inaccesible. Eventually, half of the internet is inaccesible due to running on GCP. When Google goes down, a good portion of the internet goes with it. Everyone’s eggs were in a singular basket, and when that basket broke, everyone’s eggs fell out. Now, I’m not saying that moving away from Google will instantly solve this issue. Not in the slightest. You are still running on some computer, whether it’s one you rent or one you own. However, the benefits are that you won’t be at the mercy of whatever may be happening somewhere you don’t have control over, especially if you host things personally.

Secondly, which I feel is the most important part: Google is an advertisement and search company. It’s in their best interest to know everything about you so they can give you the most relevant search results and advertisements. This can be inherently useful to getting you relevant information fast. However, as this is Google’s main source of revenue, it is really in their best interest to know as much as possible. Where can they collect this data? Anywhere you use their services: Google itself, YouTube, Android (as a platform itself, Google is baked quite deep in there), Gmail (all that online shopping you did? Google knows what you bought now!), Chrome (ever notice it absoutely begs you to sign in, sometimes signing you in browser-wide without your consent as soon as you sign into any Google service? Well, I wonder why. Maybe it has something to do with this or this?). Anything you do under Google’s eye can be used as more data points against you. Not only that, but websites serving Google Ads will also contribute to data collection, though to a lesser extent with tracking cookies being mostly eradicated very soon, though that doesn’t mean it’s impossible to track you still. Finally, with the advent of AI that started with the release of ChatGPT (not exactly, but it definitely came to the forefront with it) I am begining to have more and more reservations about storing all of my data (like photos and emails especially) with companies that are pushing to generate better and better AI models for themselves, because I wonder where the training data comes from…..

I thought to myself “Why do I need to give up all of my privacy?” Is it really worth the extra effort to run my own services to keep a portion of my privacy? To me, it really is. It’s more for the principle of the matter to me. I should be able to choose who gets my data and what I give them. No more, no less. So, how exactly did I get Google (mostly) out of my life? It really just started with wanting a solution to host my photos and grew from there.

Photo hosting

Around the end of 2020, Google announced the end of Google Photos’ free tier to be June 1, 2021. What this meant was the Google Photos backup that I had running on my phone in the background for the better part of 3-4 years at this point was about to start counting against my Google account’s storage, which is 15GB. 15GB doesn’t really go too far anymore, especially when you use Google Drive to store a random assortment of files as well, because it was “so convenient.” I really didn’t feel like paying Google and compressing my images, and with the falling costs of HDD based storage, I figured I would try to do it myself and see what happens.

In comes Photoprism, a FOSS image hosting solution that fills in the niceties of Google Photos quite nicely. Album creation, face detection, location based image searching, and live photo previews, to name a few features. They have a docker image, so it was easy enough to get running on my (measly) hardware, which was a Raspberry PI 4 (4GB model) at the time. So, I downloaded the Docker image, got my reverse-proxy set up to properly redirect the traffic to the instance, and downloaded my entire Google Photos library. Once I had it downloaded, I simply took the zip file, put it on my server, and extracted it to an ingestion folder as specified by Photoprism.

Once it was all ready, I started the Docker container and just let it churn through my photos, which admittedly took the longest amount of time (about 8-9 hours). However, this is a one time task (at this scale) that had to categorize, run facial and object recognition, and generate metadata for quite a few photos (I do not have an exact number anymore, but it was probably close to 4000 photos/videos.) Now I have auto-backups set up through Photosync that uploads photos during the night as my phone charges to Photoprism’s WebDAV endpoint. Over the past 2 (almost 3!!!) years, this has been working great!

Expanding from there

Ever since my self-hosting journey began, almost everything I use has been questioned with “Can I replace this?” Going back to “putting all of your eggs into one basket,” per se. A lot has happened since I started writing this article, however some of the biggest in recent memory has been what has hapened with Twitter and what is happening with Reddit. Let’s take a look at Reddit, for example. There is currently (as of June 13, 2023) a blackout of over 8000 subreddits, where these communities are completely inaccessible. Now, this is happening to make a point, which I wholeheartedly agree with. Third party apps shouldn’t have been priced out, Reddit should have communicated better, they are shooting themselves in the foot, the whole nine yards. However, as I stated on my Mastodon account the day before, I’m at a bit of a crossroads about the whole thing. All of that information is gone, some of it permanently (with the exception of archives, which are way harder to discover), which is… unfortunate, to say the least. It’s sad that information is just gone and so many searches are going to return way less relevant information than they once would. Maybe we shouldn’t rely on a company to host our conversations (while also trying their hardest to monetize us as much as possible at the same time?) That’s what self-hosting comes down to in my opinion: wanting no one but yourself to be responsible for the data you choose to release (or not) onto the internet. That way, if anything happens, you know who you can point the finger towards at least. :)

Conclusion

Hopefully this article wasn’t too all over the place, being written over the span of two years and all. I had a lot of topics I wanted to at least touch on but not a whole lot of “flow” I wanted to include; wanted to avoid writing more of a book than I already have. Next article in the series this summer will probably talking about setting up Mastodon, the Fediverse in general, my thoughts on decentralized services and the like. Until then, I will see you all next time!