I really wanted to love Matrix but it just doesn't cut it yet. I hope it gets there in the coming years, rather than go the way of XMPP, but atm, I can no longe

I really wanted to love Matrix but it just doesn't cut it yet. I hope it gets there in the coming years, rather than go the way of XMPP, but atm, I can no longer in good conscience try to convince people to give Matrix a shot, if it fails at covering the most basic use case: reliable communication. There aren't as many clients as you think, and the majority of them are quite bad. The only client with any firepower behind it is Element, and the UX is really not great there, and there are a lot of critical bugs that have been open for years, with nobody with enough knowledge of the codebase able to go in and fix these (especially with E2E encryption). As far as I can tell, their $30 mil raise has gone into focusing on enterprise instead. Matrix bridges are buggy and decaying because the main ones are maintained by basically one guy who's now focusing on building proprietary products around his work (fair enough). This is irrespective of the whole federation side, which is also buggy, and until the setup/maintenance cost of running your own homeserver (pure P2P federation and homeserver-agnostic identities have been TODO for years!), it's not gonna catch on. Even today, Matrix is hardly decentralised -- compare the matrix.org homeserver with literally the sum of all other homeservers and you'll see.