Okay, just wrote an Ansible Playbook to roll out the workarounds, including call audio, which I just got to work, I’ve also tried to get Bluetooth working, but the udev rule is still broken.

https://git.erebion.eu/forgejo/erebion/sargo-temp

There are also some hastily written steps for building an image. :)

Let me know what things are unclear, I will improve this.

EDIT: Only tried using call audio via headphone jack and did not notice that internal speaker and mic currently do nothing, so I have to amend my to do list by just another bug. :D

EDIT 2: The mic patch has not yet been pushed by the kernel dev and the ALSA config needs changes, but then audio calls with built-in speaker and mic will work. :)

Started writing, but then got sidetracked… by making audio calls work. :)

Will post the link here soon, just want to verify there are not too many dumb mistakes in it, do not want to fruatrate anyone with terrible docs. :D

Das mit den Metadaten ist bekannt. In Zukunft soll die User ID nicht mehr Primary Identifier sein. Da das Protokoll dann aber stark verändert werden muss wird das nicht in den nächsten Wochen passieren. Da wird jetzt schon eine Weile an möglichen Updates fürs Protokoll geschraubt.

Seit dieser Woche hat Synapse experimentellen Support für Sliding Sync, in Kürze sollte es dann einfach da sein und dann wird es auch deutlich schneller.

Ansonsten ist ein häufiges Problem, dass Leute auf matrix.org sind, und der Server überlastet ist.

Und wenn ihr eure eigenen Server auf einer Version von Postgres 15.x habt, dann probiert mal auf ein neueres 15.x zu aktualisieren, da war irgendwas drin kaputt, bin allerdings nicht ganz sicher ob das nur ARM betraf.

Zumindest ist Matrix eher dann sehr langsam wenn was kaputt ist, sonst merke ich da keinen Unterschied zu Signal.

You won’t have to wait for very long.

Let me know if you’d like a writeup of the way I currently build the image out of the linked sources and what files (temporarily) need manual fixes, in case you want to try it out. Could still take a while to get the kernel and the call audio daemon ready for Mobian and to get it included, but other than that… Pretty close ^_

If you exclude the time your PC does compiling and works for you, then it is probably 20 minutes work to get an image currently. :)

I also have an Ansible role which does the few fixes I need for now, like that one systemd unit. I could publish that as well in case others want to be alpha testers and tell me about any issues they notice.

I should probably also create a repo with a temporary issue tracker until there are official images. 🤔

Thanks! I will definitely be able to. I could already have it as my daily driver using VoIP, until call audio works, but I know I could already compile the daemon myself and it will just work. :)

But as I want others to be able to use it at some point, I will delay using it as a daily driver and instead make sure the few remaining issues get sorted out.

Much better than my Pinephone, lol

I cannot complain. I have not used it for a whole day yet as I am mostly fiddling with the kernel packaging and I want to get call audio working before I use it as my daily driver, but I have no reason to believe it would not last the whole day. No issues here. Probably longer battery life than with Android. It does not constantly talk to Google, after all. xD

It is, but you still need the app running somewhere.

I’ve posted a follow-up, it looks much better now: https://lemmy.sdf.org/post/19600671

Yes, but even the app could be avoided altogether with the other authentication method implemented. :)

Exactly that.

It does not support the app’s authentication yet, so there would be a noticeable difference with that implemented.

Will have to post a proper update soon-ish, but for now I’ll just say that the device completely* works and I just have to get the kernel package into Mobian (until patches are upstream), for that get some Kernel packaging stuff right (it already builds, but is not yet the quality Debian people expect) and then create a few smaller MRs at projects which Mobian is built upon.

*everything except some sensors and some patches which exist but aren’t yet pushed to git…

And hopefully I can save a few devices from getting trashed. :)

Made huge amounts of progress since I posted this! Should really post a follow-up at some point. :D

Sending all the voice messages I get to some server? Uhm, no.

Great idea in theory, though.

A bunch of marketing blah blah, I wonder what they’re up to.

Do you get fined if you are unlucky to have no harvest and cannot bring anything back?

I don’t know what would fit your needs, but Signal does not require Play Services. And even if those are present, it does not leak data to Google. Other than “Signal is installed” and “You get a push message”, Signal does not put your messages into the notifications. Instead Signal connects to the Signal servers and then gets the encrypted messages from there and only then decrypts.

Even if you have Play Services installed, you can force it to use a background connection inatead, if you disable Play Services before installing Signal, it wall automatically fall back to it.

If you want a version without Play Services libraries, you could use Molly, a hardened version of Signal, which is available in a version without those libraries.

Molly even allows linking phones as secondary devices, not just desktops.

Matrix is not the right protocol for staying anonymous. There’s way too much unprotected metadata.

You might be able to mitigate that somewhat by using an instance that is accessible via TOR and being careful who you communicate with, depending on threat models and so on.

But if you want to communicate anonymously and not leak meta data… Probably not what you are looking for.

• user names cannot be changed
• never delete your database, have backups, all the state is in there, if you lose your DB, you essentially have to set up a new server: Your server will say you are not in a room while others say you are, so you cannot join. Always keep your DB backed up.