Yeah, no. I run both a Matrix server (Synapse) and an XMPP server (Prosody). It is completely obvious when Synapse is running on my server. I don't need ps or systemctl status, I can tell if Synapse is running just by looking at the bars at the top of htop.
Meanwhile my XMPP stuff is as invisible as any other system service.
Re-reading my posts, I guess I could be more specific. Y'all seem like you have great technical vision but are having a hard time with success, growth, and revenue. While I can appreciate some technical bits of Matrix, I would never dare to think I could do better than anyone at those real challenges. It hurts to read about y'all struggling for dev funds, in spite of success. And I feel like that probably weighs on you, it would weigh on me. Godspeed, good luck, I look forward to tomorrow's TWIM :).
Synapse is very stable these days, and relatively resource efficient. Matrix.org (both the server, website, spec and foundation) is in a good place too, as is (from my biased pov as project lead) the project in general.
> This is a huge "oopsie" to put it nicely
We simply don't have the people to do a good job of maintaining the Libera bridge. We tried to jump forwards on it with persistent connections to improve the experience on the Libera side; this caused a tonne of instability; we couldn't move back (given it would have also been seen as a backwards step); we didn't have the manpower to fix it rapidly either; so this is not remotely our preferred outcome. We ran the bridge for almost 8 years; it wasn't perfect; it wasn't a total disaster either; but in the end we overstretched ourselves and Libera pulled the plug.
> It feels like a massive loss for matrix (bad first impressions, negative impact on network effects, etc) and will undoubtedly be something that matrix detractors will point to for years.
Matrix has around 5M native active users based on the phonehome stats, of which ~22k were using the libera bridge. I agree that it's awful for early adopter geeks who love IRC, and I despair that it has happened. But I don't think the sky is falling either.
> Element X has been pleasantly satisfying lately. I'm curious if the team is deliberately taking a different strategy.
Yes, it's a very different strategy. Element isn't working on the legacy mobile apps at all any more (other than for security updates), and rather than being spread too thin over too many projects and codebases (Element Web, Element iOS & Element Android all had zero code in common other than libolm) instead the intention is for Element to focus on one primary codebase (matrix-rust-sdk + Element X and Element Web 'R') and nail it. Meanwhile we don't have the manpower to do a good job of other projects (e.g. Libera bridge, P2P Matrix, Third Room etc.) and so they are on hold for now.
I agree that with 20:20 hindsight this hasn't been the most efficient route to get to where we are today.