![](/static/253f0d9/assets/icons/icon-96x96.png)
![](https://lemmy.ml/pictrs/image/ucPeLo62DS.png)
I haven’t tested it, but did you look at Damselfly? The documentation seems to suggest you can do it: https://github.com/Webreaper/Damselfly/blob/master/docs/Multi-user.md
I haven’t tested it, but did you look at Damselfly? The documentation seems to suggest you can do it: https://github.com/Webreaper/Damselfly/blob/master/docs/Multi-user.md
Hard to argue with Intel, but I run one of the asrock j3455 boards (with a full PCIe slot and 2 SATA ports) and powershell is reporting OSTotalVisibleMemorySize of 12228504.
Pretty sure I am running a j3455 with 12GB.
edit> Confirmed
No major issues with the upgrade to 14, yet. On a pixel 7a. Not that I spend much time in the play store, but that design has become less pleasant over time.
The standalone converse app was problematic when I tried it last. Also, there was a summer of code attempt at bringing jingle a/v sessions to converse, but it was never completed and nobody seems to have picked it up.
There is also Cheogram (conversations fork), which is actively developed/tweaked by the jmp.chat folks - very nice. Also Snikket (conversations fork) that is themed and tweaked to use with a snikket server, but it happily works with other servers.
Another interesting tidbit. Chromebooks integrate the Android runtime to run play store apps. Windows 11 is also kinda/sorta shipping an Android runtime, but not by default. You can also spin up an Android runtime on Linux. I tested the snikket android app on Windows 11 and ChromeOS - works perfectly. So, I suspect all conversations forks can run across Android, Windows, ChromeOS, and Linux platforms - pretty neat. Doesn’t solve the iOS gap and getting the runtimes going could use polish on Windows and Linux. And nothing against the other desktop apps in development, but the ability to essentially run the Android app against most major environments makes me want to contribute to that code base (if I had any ability to develop for android, that is).
If a sizable chunk of the Reddit community can move to an alpha/beta grade link aggregator platform that can’t handle the load because we believe in the decentralization or the instance’s mission or the overall concept of federation, why is it too late for the community to re-adopt a mature messaging platform that mirrors those ambitions?
xmpp has clients doing voice and video - it has for years. It is p2p and falls over in some nat to nat situations, which is where stun/turn come in on the server. Check out jmp.chat - they built a voip phone service using xmpp clients.
I would like to hear more about what you’re doing / how you have it set up. I’ve used xmpp to relay messages from home automation stuff - which usually involves piping something to a script calling a library.
Omemo is double ratchet and my messages sync to multiple devices. New device can’t read old messages sent before exchanging keys with the other clients.
I haven’t used Matrix for messaging, so take this with a grain of salt. But xmpp servers and clients seem to be lighter on resources. Matrix has more capabilities for large groups.
Google’s messaging play has only gotten worse since then. Oh well.
I use it for OMEMO encrypted family messaging and image transfer (snikket). Very fast messaging, lightweight server, and the A/V works quite well. Biggest issue, imo, is the lack of a great iOS client - not a judgement on the developers, I think that’s just the reality of developing on iOS. But an iOS client that works as seamlessly as Conversations would go a long way to regaining lost traction.
Mature web framework and highly productive language vs less mature framework and emerging language. Personally, I think Rust is the more surprising pick than PHP for this application. A link aggregator is a forum with some frills. Not to mention half of the activitypub implementations that I know of have been in PHP.
The average person isn’t going to delve into the nuance of open source project structure. If I wanted to support the jellyfin ecosystem, I would probably expect that donating to the jellyfin project is sufficient.