Andrew Melder

With recent world news making many of us rethink our relationship with technology, specifically our relationship to U.S technology companies that run the most popular services being used; I decided to take the plunge and try moving to GrapheneOS on my Google Pixel 9 Pro.

The goal

Despite the common reason for those looking to get a custom deGoogled ROM, I’m not looking to eliminate Google from my online activity. Instead, I want to reduce my reliance on their services wherever I can.

The primary way of doing this was moving my email account away from Google, and all the services attached to it. While that is still in progress, I’ve made enough of a head start that I thought I could make it work. But was I a bit too keen and would that compromise the migration?

Barriers I knew heading in

Of course, going from having a lot of your services centred around Google makes leaving the ecosystem somewhat problematic with adjustments needed. For myself, in regard to moving to a deGoogled ROM, there were two main concerns:

1. There’s apps I need from Google Play (banking, work apps, Google apps etc). While GrapheneOS has a sandboxed Google Play option to allow full access to the store and its functions, will there be any issues with the services I need?

2. I have a Pixel Watch. I actually quite like it, something I wasn’t expecting coming from the Apple Watch ecosystem prior. Given the Pixel Watch relies heavily on Google Services, will there be any issues and can I even get the watch synced and working in a usable state?

Installation and the first 24 hours: What have I done?

The web based installation itself went as smoothly as you can imagine, in terms of installing a custom ROM onto an Android phone. If you’ve done this in the past (even if it was many years ago), you would notice that this is a far nicer and simpler experience. It’s still not a process I would tell any non-technical person to do, I don’t think any custom ROM process can be; however it is about is seamless as it can get.

What you are greeted with once it is complete though is bare bones Android, an App Store with a handful of apps and no onboarding for those not of a technical mindset. This is not an experience for the faint hearted or for those not willing to dig into things and work it out themselves, or at least with a little online help from guides out there.

I proceeded to install a few items, then also look at some open source alternatives through F-Droid and Obtainium. I eventually got to installing the sandboxed Google Play service through Graphene’s App Store app, logged into my Google account.

My main goal for the Google Apps at this stage was to get my Pixel Watch synced and running. With the Watch app installed, I setup the Watch with the phone itself without too many issues. However it was when I started to use the watch that I realised just how dependent on Google connectivity the watch is.

The first issue I noticed was that my Do Not Disturb setting was not syncing between the watch and phone. After playing around with permissions settings, I realised from the Watch app that this sync requires Google’s Digital Wellbeing service; which is both not installed nor can it be installed from the Play Store (not compatible based on the Play Store listing). I used this a fair bit but now I’ll need to manage these separately (I find setting it on the phone prevents unwanted calls and notifications coming to the watch anyway).

The second issue was the barrage of Fitbit notifications from the watch. The watch app itself is able to update directly via the Google Play store without needing to install the Fitbit app, however trying to access any of the Fitbit stuff on the watch will require downloading and setting up the phone app. I only care about the step count anyway, which is still visible by a tile or complications on the watch without needing to install the app; so I still haven’t bothered with that.

The third issue was not getting ring notifications on my watch for incoming calls until after I missed the call. Trying to stick with Graphene’s stock phone and messages apps was the plan, and messages came through fine and I could even respond from the watch without issues. However, calls would not show up on the watch until the missed call notification came through.

It was at this point, I started to wonder if this was all worth it and thought I might be moving back to stock within 24 hours.

This is a big issue for myself. The primary reason I have a watch is for call notifications (I had some important missed calls over the years from my wife when I didn’t have a watch, so now she insists on it). Installing Google’s own phone app fixed this and instantly enabled the ability to take calls from the watch too. I’m not too concerned about using the Google phone app as it has very good spam call filtering and is generally nicer to use than stock.

If you use RCS, you will notice the stock messages app doesn’t support it either for privacy reasons. While I don’t use it so don’t have an issue with this, others may want to install Google’s own messages app as well.

Working through the kinks and getting back to normal operations

It not all problems though as there has been some things on the watch that are working which I didn’t expect to on a custom ROM with restricted Google access.

The calendar app gets its info directly from the cloud, so you can see your Google calendar and tasks entries directly without needing to install or sync these to the phone. Pretty useful for me as we still have a family calendar in Google (and I’m not forcing my wife to change again after migrating away from iCloud).

The same applies to Keep notes, where I would use Google’s Assistant to add and update my shopping list in Keep. With the Keep tile also working, updating and ticking off my shopping list is exactly the same as it was prior. Speaking of Assistant while you do need to install the Google app to configure it on the watch initially, once this is done the Watch assistant runs locally and the Google app can be uninstalled from the phone.

Google Pay was something I definitely wasn’t expecting to work. While I knew that tap-to-pay though the phone was not possible due to security restrictions, I did install the Wallet app for all my other passes already saved to the Google account. I was pleasantly surprised when I was able to setup my payment card in Google Pay for the watch and use it successfully. In hindsight, as long as the watch can setup the card, there’s no reason it shouldn’t work as the Pixel Watch as its software is still stock; but it was an added benefit I didn’t expect. Note: In order for the Google Wallet app not to crash on loading, you must allow Phone and SMS permissions for Google Play Services. This might be a privacy red flag for some.

My mistake: trying to look for Google alternatives straight away

While I came into GrapheneOS to start my journey away from Google, I got caught up in trying to do things without Google’s oversight straight away. As a result, I almost got too frustrated trying to work out the issues I was encountering and considered going back to Google’s OS instead.

I was reminded by Graphene’s Mastodon account itself that it’s fine to start with all the Google services setup and work your way backwards from there.

Next steps in the journey

I’m still working through the pros and cons of routing Geolocation access requests through Graphene’s system instead of enabling the permission through Play Services and using Google’s Location Accuracy functionality. It is definitely slower to get a location and not as accurate, which mainly impacts using Google Maps.

My next step is to re-establish my work profile and MFA apps for use at work. I have my old iPhone 13 setup as a backup for now with my work stuff on it, so it’s not a critical thing to get working ASAP.

Why bother if I’m just using Google services anyway?

There’s probably many who are asking this question. I mean if I’m just installing and enabling Google services anyway, why go through all of this trouble?

Well the main thing is a combination of agency and privacy of my device over Google. There was a recent discovery of Android System SafetyCore service secretly installed onto Android devices to assist with on device scanning of images for unwanted content; which raised the same privacy concerns as Apple did a few years prior.

The other thing is just the volume of services Google installs on Android phones and how often it sends data back to Google. As per the screenshots, there’s just a bunch of stuff I don’t want or need, but I don’t have a choice with the stock OS. Also, when I did have location enabled for Play Services, it was curious to see just how often it accessed my location even when I was doing an activity not related to location at all (a.k.a opening my Mastodon app).

Combined with the security hardening the GrapheneOS team have done which goes beyond what Google have; even if I continue to use Google services themselves, it seems to be a benefit to run this more locked down OS.

But where does an e-ink smartwatch exist in a modern market now packed with competition?

Eric Migicovsky, the founder of Pebble smartwatches, is bringing the company back and planning a new watch after Google agreed to open source the original OS.

As a long time fan and owner of the original Pebble and smartwatches in general, I think this is fantastic news. As a gadget lover, I’ve also signed up right away and will (assuming it isn’t a stupid cost) look to purchase the new model if I can.

I think more options in any category is great for the consumer but I have to admit that as I thought about this and signed up, I looked down at my perfectly capable Pixel Watch 2 (as well as thinking about my previous Apple Watch experience) and started wondering where this new effort fits in the market.

A niche product, but now with appropriately small ambitions

The original Pebble was somewhat of a market leader at the time. The technology needed to enable the functionality of today’s smartwatches, with bright touchscreen displays, sensors for health tracking and speakers and mics for on-the-go calls, just didn’t exist at the time without strapping a literal smartwatch to your wrist.

While they definitely weren’t the first to try the concept (companies like Sony Ericsson would try smartwatches almost a decade prior), they were one of the first to gather a more widespread appeal among tech circles and others with their Kickstarter campaign and line of e-ink watches.

What they offered was simple; e-ink screens that could be read almost anywhere, simple software that was easy to navigate while providing basic smartwatch functionality, amazing battery life, compact designs that didn’t look ridiculous on your wrist (early Android Wear enthusiasts would know the wrist pain).

They were initially successful, gathering a loyal fan base; many of whom continue to even use the watches to this day. However (as Eric has outlined himself) the second model, the Pebble Time, was ambitious with its growth targets & over project deadlines, with a lack of clarity around its value proposition for the mass market (first productivity, then fitness).

In short they grew too fast, over-estimated the demand, ended up overstocked with unsold inventory and started to lose money.

It appears that Eric and the team are getting back to the basics with the relaunch, sticking with a product they want to use themselves and keep the ambitions small to the similar group of initial user base that wanted the same simplicity in a smartwatch.

Keeping things simple

The really interesting revelation in Eric’s note is that they really are starting where Pebble ended, planning to build the newest watch with the same specs as the last platform.

Along with restarting work on the now open source Pebble OS platform I’m hopeful this approach will not only keep things profitable for the rebuilt company, but also keep the costs of the new hardware down to something many people are tempted to try out.

In my mind, the reality is that the smartwatch market itself is starting to mature. People are still happy using Apple Watches from a number of years ago as the hardware and software still work fine. Wear OS (or whatever it’s called now) has had a few more speed bumps but is now also maturing, even if the ecosystem is mainly Samsung and Google presently.

As a result, decent second hand watches are good value, clearance models when a new watch is released are almost a steal, and entry level new watches are more than capable for most people.

As such, even to its niche market, I think the new Pebble has to be cheap. Its existing advantages are minimal in the mature market, and users are no longer adverse to charging their watches nightly as they do their phones; to the point where I think the advantage of a watch that last multiple days is pretty irrelevant to many.

But the encouraging thing is that the new Pebble isn’t interested in the requirements of the many; just the small group of like minded individuals who want what they are offering.