Most active commenters
  • palata(8)
  • BeetleB(3)
  • ihatehn(3)

←back to thread

Leaving Gmail for Mailbox.org

(giuliomagnifico.blog)
351 points giuliomagnifico | 43 comments | | HN request time: 0.001s | source | bottom
Show context
TranquilMarmot ◴[] No.44988014[source]
I spent the past month "de-Googling" my life after I saw a notice in my Gmail inbox that it was 20 years old. I took a step back and realized just how invested into the Google ecosystem I was. Gmail, Calendar, Docs, Drive, Maps, Keep, Photos, YouTube, FitBit, Android. Basically my entire digital life. My goal was more diversifying than security/privacy, but security/privacy is a really nice bonus.

I ended up going with Proton because they had a good solution for mail, calendar, and drive which I was looking to replace. I set up my custom domain to point to it and have my Gmail forwarding to it - any time I get an email to the old Gmail address I go change it on the website or delete the account altogether.

For Google Docs / Keep, I switched over to Obsidian and pay for the sync there. It's a great replacement for my main use case of Docs / Keep which is just a dumping ground for ideas.

For Google Photos, I now self-host Immich in Hetzner on a VPS with a 1TB storage box mounted via SSHFS. I use Tailscale to connect to it. It took a few days to use Google Takeout + immich-go to upload all the photos (~300GB of data) but it's working really well now. Only costs $10/mo for the VPS and 1TB of storage.

Android I think I'll be stuck on - I have a Pixel 8 Pro that technically supports Graphene but there are too many trade-offs there. Next time I need a new phone I'll take a serious look at Fairphone but I think the Pixel 8 Pro should last a few more years.

My FitBit Versa is really old and starting to die - I ordered one of the new Pebble watches and am patiently waiting for it to ship!

YouTube I'm stuck on because that's where the content is. I have yet to find a suitable replacement for Google Maps - OpenStreetMap is still really hard to use and gives bad directions.

replies(20): >>44988286 #>>44988592 #>>44988765 #>>44989953 #>>44990098 #>>44990152 #>>44990278 #>>44990388 #>>44990401 #>>44990425 #>>44992000 #>>44993497 #>>44993728 #>>44993924 #>>44993946 #>>44994247 #>>44994386 #>>44997433 #>>44998116 #>>45001272 #
1. palata ◴[] No.44990388[source]
> supports Graphene but there are too many trade-offs there

What are the tradeoffs? I have been following GrapheneOS for a while, and it doesn't seem like there are many tradeoffs.

> OpenStreetMap is still really hard to use and gives bad directions.

OpenStreetMap is a database, and most commercial services that are not Google use it. E.g. Uber or Lyft.

You just need to find an app that you like. CoMaps is nice, OSMAnd has a lot of feature but the UX is harder. And of course you can contribute to OSM and make it even better than it is! You'll see it's a great community!

replies(5): >>44990590 #>>44990913 #>>44994958 #>>44997375 #>>45005398 #
2. nine_k ◴[] No.44990590[source]
Can you use GrapheneOS with your bank app? With a digital wallet for NFC cards? With Uber or Lyft? (Asking seriously, not rhetorically.)
replies(8): >>44990884 #>>44990889 #>>44990926 #>>44990938 #>>44992858 #>>44994924 #>>44995398 #>>44996543 #
3. octo888 ◴[] No.44990884[source]
Contactless payments is the the big one that doesn't work and probably won't. You can do in app payments via Google pay though

Many banking apps work fine though not all.

replies(1): >>44994774 #
4. callahad ◴[] No.44990889[source]
My understanding from looking into this two years ago is that it's hit or miss for banks (depending on if they opt into device attestation stuff), no for NFC / Google Wallet, and yes for Uber / Lyft.

Apparently the common workaround for the Google Wallet stuff is to pair a GrapheneOS phone with a stock Android smartwatch.

Edit: Here's some additional information on banking apps: https://privsec.dev/posts/android/banking-applications-compa...

Apparently the common recommendation these days is to use Curve Pay as a virtual card provider on GrapheneOS, which can then route to arbitrary underlying cards. And evidently Google Wallet does work for things that aren't payment cards (airline tickets, transit passes, etc.) on GrapheneOS.

5. BeetleB ◴[] No.44990913[source]
Someone showed me OSMAnd recently while we were hiking. I installed it as soon as I got home. Great for hiking.

Then last week I used it for navigation (on a phone with no SIM card).

Absolutely. Terrible.

Worst navigation app I've seen. Told me to make a turn at an intersection that did not allow turns. Then at another intersection, it told me to "Turn left", but the display clearly showed it going straight. I'm guessing that the straight road probably is angled 1 degree or something at the intersection and the app was viewing that as a turn.

replies(5): >>44992229 #>>44994026 #>>44995340 #>>45004411 #>>45016049 #
6. BeetleB ◴[] No.44990926[source]
Yes, these would be my concerns as well. In the past, I would install custom ROMs. Then I stopped doing that and would only root my device. But of late, way, way too many apps refuse to work if rooted (apps that used to be fine with it before).

Now I just accept life as it is.

replies(1): >>44998898 #
7. jackthetab ◴[] No.44990938[source]
This is a question that I rarely see answered but would love to know as well.
replies(1): >>44998910 #
8. neilv ◴[] No.44992229[source]
For an open source Android app for OpenStreetMap data, I like Organic Maps, and it normally works great with locally-cached maps. I've had better luck with it than with Google Maps or Apple Maps on phones.

(Though, I should mention that twice in the last year I've had Organic Maps become hopelessly confused about where I was, and where I should go. Both times, it had gotten a good GPS location, but then got confused while being out for an extended period of time, like maybe it was dead-reckoning only after that initial lock.)

replies(2): >>44993752 #>>44994927 #
9. litmus-pit-git ◴[] No.44992858[source]
My friend uses a pretty hardened (as per him; I didn't indulge him when he wanted to give me the gory details) Graphene setup on his few years old Pixel.

Bank apps - as per him none work. Uber (no Lyft here; other taxi apps) work flawless. Payment apps, he said is a coin toss. On his phone even WhatsApp doesn't work. He anyway prefers Signal (which prob. nobody else uses in his circle except maybe me who has it installed on a secondary phone) or plain SMS. Basically most of the "normal" apps that add integrity checks don't work but he is fine with that.

replies(1): >>44994688 #
10. crinkly ◴[] No.44993752{3}[source]
+1 for organic maps. Have used it hiking and travelling all over the world. Never had any issues with it.

Not had any GPS problems other than that time I was in an area where it was being jammed. Bloody Russians.

11. incone123 ◴[] No.44994026[source]
I get similar navigation issues with Google maps. I still use Google maps for driving because the live traffic is important to me, but other posts on here mention other apps with live traffic so I'll give them a try.
replies(1): >>44997924 #
12. hexfish ◴[] No.44994688{3}[source]
Re: the bank apps: that really depends on the bank and the country. I live in a eu-west country and there are afaik no apps that do not run on Graphene (which did suprise me I must admit).

Whatsapp can work if you use sandboxed Google Play (I still use a Google account, I just don't want gplay to have effectively root on my device).

Depending on the level of integrity check your app might just work. Gory details: https://grapheneos.org/articles/attestation-compatibility-gu...

And like others said: no contactless payment, but I dont use that personally anyway.

13. amaccuish ◴[] No.44994774{3}[source]
Luckily my German bank (Volksbank) has its own NFC app on Android. Much maligned years ago when it was announced (why can't they just let me use Google Pay??), I at least have come to the conclusion that it has granted me a freedom that Google does not.
replies(1): >>45022501 #
14. mlry ◴[] No.44994924[source]
I use Graphene but with Google play store app. Here in Europe my banking apps and 2fa apps (SecureGo) work flawlessly. NFC cards work with PassAndroid and FOSSwallet, both from Fdroid. I've had issues installing rather new games via the play store, but most often it takes a couple of tries or a waiting period to work in the end.
15. Freak_NL ◴[] No.44994927{3}[source]
Regarding Organic Maps: I would recommend keeping tabs on what is happening there since this year. They seem to be having significant governance issues.

https://community.openstreetmap.org/t/organic-maps-open-lett...

Short story: forget Organic Maps, use successor CoMaps or competitor OsmAnd.

https://www.comaps.app/about-us/

replies(2): >>44995454 #>>44997114 #
16. mlry ◴[] No.44994958[source]
I use https://brouter.de/brouter-web on my laptop. Someone told me that you can use brouter as the nav engine for Osmand and thus greatly improve speed and accuracy for navigation, but I have not yet tried this.

And I recently installed GMaps WV from Fdroid as a wrapper for Google Maps. It gives current traffic information but I don't really know if it is even close to gmaps.

17. rpdillon ◴[] No.44995340[source]
This is a really interesting feedback. I've used OSMand for maybe five years, and never had issues like you're describing. I've always felt that the search was absolutely awful, so I used Google Maps for that and then put the points of interest into my map. Nevertheless, I find the display particularly dense and confusing to configure, and so I also have been using Organic Maps lately, which may provide a simplified experience that's a bit more polished.

I wonder if there was some issue with the map data in the area you were driving in that led to the issues you experienced. I've used OSMand in Belize, Mexico, California, Connecticut, New York, New Hampshire, Vermont, and Maine and had a good experience, especially with the offline maps.

18. snapplebobapple ◴[] No.44995398[source]
I use it with my bank app ymmv
19. rpdillon ◴[] No.44995454{4}[source]
I was vaguely aware of this drama but hadn't looked into it. After reading through your link, I've switched over to Comaps. I don't like lack of transparency in community driven projects. Appreciate the flag!
20. fak3r ◴[] No.44996543[source]
I've been running GrapheneOS for over a year and have had zero issues with 3 different banks, and all credit cards. I'm sure there are issues with some banks, but I've never seen them. I don't use Google Wallet, and never wanted to so if that's a consideration...
21. neilv ◴[] No.44997114{4}[source]
Thank you, that's interesting and a concern.

Do you have pointers to information about the governance and legitimacy of CoMaps?

(I see a mention that it's non-profit, but no statement about what kind of non-profit, not even on the donate page where that info is customary and relevant for US tax reasons. Also, I see no mention of who's who, nor how they operate.)

The closest I find is this:

https://www.comaps.app/support/what-is-the-comaps-history/

> As a result of the issues not being resolved, in April 2025, the community of former Organic Maps contributors created the CoMaps project, based on the Organic Maps open-source code.

If what that sounds like is true (that it does represent the community of contributors), it still will be important to have safeguards against someone taking over the project.

Or, if what that sounds like isn't true, that could be bad.

One matter that will have to be resolved with governance (if it hasn't already), is that there's what looks like an allegation that the CoMaps project is already tainted with code to which is expressly doesn't have license:

https://codeberg.org/comaps/comaps/pulls/1039#issuecomment-6...

A concern is that a funded commercial competitor could bankrupt a less-funded volunteer project with lawyer fees just arguing the merits of that.

replies(4): >>44997182 #>>45011866 #>>45023431 #>>45037047 #
22. Freak_NL ◴[] No.44997182{5}[source]
Nope, all I know is what I picked up in the OpenStreetMap community, like that thread on the forums, and this open letter:

https://www.comaps.app/news/2025-04-16/1/

replies(1): >>45011491 #
23. TranquilMarmot ◴[] No.44997375[source]
> What are the tradeoffs?

As others point out, my main worry is about banking and NFC. I use NFC payments on my phone a lot, especially for the bus. Getting an Android Smartwatch just for that kind of defeats de-Googling haha.

I will probably try out Graphene at some point but that seems like a multi-day project to get it set up, find the tradeoffs, determine if they're worth it, and then potentially switch back to Android.

I also worry about the future of Graphene with AOSP going more closed/private: https://www.androidauthority.com/google-android-development-...

> OSMAnd has a lot of feature but the UX is harder

OSMAnd was the one I tried and it bordered on unusable. I'll try out CoMaps, somebody else suggested Mapy.

replies(3): >>44998300 #>>44998864 #>>45022527 #
24. BeetleB ◴[] No.44997924{3}[source]
This happened close to my home on a road I've gone on almost every day. Waze never did that to me. I don't use Google Maps so I can't speak to that.
25. nabakolu ◴[] No.44998300[source]
There is also Calyx, which is also a privacy friendly Android distribution. It uses open source versions of Google Services. With my Pixel 8a on Calyx I can use NFC to pay with my Credit Card.
replies(1): >>44998882 #
26. palata ◴[] No.44998864[source]
> my main worry is about banking and NFC

This may be useful: https://privsec.dev/posts/android/banking-applications-compa...

> I also worry about the future of Graphene with AOSP going more closed/private

I don't think it's AOSP going more closed, but rather the evolution of the Pixels line. GrapheneOS wants the ability to unlock the bootloader, install their own key and relock the bootloader. Google could technically stop supporting that for the Pixels (just like many (most?) manufacturers do), but they haven't said anything about it.

GrapheneOS has mentioned talks with a major manufacturer. I wonder if it will be a Graphene phone (which sounds very tricky) or if it could be something like Huawei or Samsung offering to unlock/relock the bootloader and committing to providing the features necessary for GrapheneOS. The latter would be amazing.

> OSMAnd was the one I tried and it bordered on unusable.

I love it. But if that's your feeling, definitely try CoMaps: in terms of UX it's on the other end of the spectrum :-).

27. palata ◴[] No.44998882{3}[source]
> It uses open source versions of Google Services.

Microg, right? I don't think this is more private than the Google Services, is it? It's just open source, but it can phone home just the same?

> With my Pixel 8a on Calyx

Any reason you haven't tried GrapheneOS, given that you already have a Pixel? I have /e/OS because it's the only alternative Android that runs on my FairPhone 3, but I've come to realise that it's actually worse than the Stock Android in terms of security. I really would like to go with GrapheneOS, and I wish more manufacturers made this possible.

28. palata ◴[] No.44998898{3}[source]
Using GrapheneOS is not the same thing as rooting, though.
29. palata ◴[] No.44998910{3}[source]
Nice answer (not by me) here: https://news.ycombinator.com/item?id=44990889
30. RunningDroid ◴[] No.45004411[source]
> Told me to make a turn at an intersection that did not allow turns.

That's an OpenStreetMap bug, the intersection likely isn't marked as not allowing turns. If you put a note (OSMAnd calls them OSM Notes) on the map someone will fix it when they can.

31. jacooper ◴[] No.45005398[source]
All navigation apps that are built on osm have awful routing, particularly with public transport, they are almost useless. That's without taking the missing data into account (shops, opening times, basically zero reviews, etc..)
replies(2): >>45007184 #>>45016125 #
32. palata ◴[] No.45007184[source]
Uber and Lyft are built on OSM.
33. opensourebuild ◴[] No.45011491{6}[source]
The claim about open-source is coming from the shareholder of Organic Maps, Alex, but there is no basis to it, as the code for that repository is licensed under the Apache 2.0 license.

It appears Alex is angry about the fork and doing anything possible to spread negativity.

replies(1): >>45028757 #
34. ihatehn ◴[] No.45011866{5}[source]
Very astute! Legally the code is owned by each contributor and licensed via the DCO. Financially the project is underneath the umbrella of the Platform 6 co-op (see OpenCollective)

This is temporary though and a permanent nonprofit home is a top priority.

35. ihatehn ◴[] No.45016049[source]
On top of the OsmAnd user experience being a little rough (it can do a lot, but not gracefully) it relies on accurate OSM data underneath. So the best thing to do is at least make a Note on OSM, or edit it yourself, and mark that intersection as having a no-turns-allowed restriction.

Even Google is relying on user submissions to keep its stuff accurate these days, they just have money to pay editors and reviewers.

36. ihatehn ◴[] No.45016125[source]
It's all user submitted. It takes someone caring enough to fix it, to make it work. Check out StreetComplete, Every Door, Vespucci, and/or CoMaps to help add/fix data
37. wkat4242 ◴[] No.45022501{4}[source]
And it doesn't let Google snoop on your every transaction. I don't understand how so many people are ok with this.

Unfortunately every EU bank card uses Mastercard or visa so the Americans still get to snoop. Time for that to be changed too.

38. wkat4242 ◴[] No.45022527[source]
Osmand is great but it has a bit of a learning curve. It's much more powerful than Google maps though so all these features need a bit of navigating. It probably should have a simple mode for most users.
39. infinitesector7 ◴[] No.45023431{5}[source]
The code is open-source. biodranik is the main source of the governance issues.
40. biodranik ◴[] No.45028757{7}[source]
The code used by the fork was never published. It was stolen from a private repository and a private server, and then published/used in the fork without the authors' approval. That's a serious legal issue.

The fork also took the new website design that was developed for Organic Maps even before the Organic Maps website was updated.

Don't believe everything on the internet; there are many lies spread around.

replies(2): >>45035008 #>>45037034 #
41. roundcast ◴[] No.45035008{8}[source]
So Organic Maps is not actually open-source?
42. palata ◴[] No.45037034{8}[source]
Hmm... This says otherwise, with detailed explanations and screenshots: [1][2]

It shows that you, biodranik, removed the MIT licence from the repo with a commit saying "No MIT yet, sorry". It says that the code had been licenced as MIT since 2021. It is not clear if you own the copyright to all the contributions since 2021, and therefore it is not clear if you are legally allowed to remove the MIT licence.

It also says that the fork was made from the repo in the state it was when you removed the MIT licence. Therefore it is a fork of an MIT project by someone who had legal access to the MIT-licenced code: it's legal.

> Don't believe everything on the internet; there are many lies spread around.

You don't say :-).

I was not entirely sure about the CoMaps vs Organic Maps situation, but this very comment of yours clearly favours CoMaps IMO. Or did I misunderstand something?

[1]: https://github.com/orgs/organicmaps/discussions/9837

[2]: https://web.archive.org/web/20250815050441/https://github.co...

43. palata ◴[] No.45037047{5}[source]
I think this is relevant:

https://github.com/orgs/organicmaps/discussions/9837

https://web.archive.org/web/20250815050441/https://github.co...

In the discussion you link, biodranik tries to say that it was illegal for those with access to the repo to dispose of the open-source code because that open-source code had not be published. I don't think this is right:

Open-source does not mean at all that the code has to be public. It can be private to a community. But whoever has access to that code is allowed, under the open-source licence, to dispose of it under the terms of the licence. My understanding is that this is what happened with the fork of Organic Maps. And it is pretty clearly explained in the codeberg discussion you linked.

> an allegation that the CoMaps project is already tainted with code to which is expressly doesn't have license

I don't think it says that CoMaps doesn't have the licence. It says that whoever forked the codebase did not have permission by some of the founders of Organic Maps. That's very different: the licence doesn't say that you must ask permission to the founders; it says that the code can be used under the terms of the licence. A fork part of those terms.