I think this is a pretty fair summary. I've been using matrix since before the patreon was started and it's gotten a whole lot better for sure, but there is still no single client that does everything. Element X is right out since it doesn't support spaces yet. SchildiChat next comes close for me, but it can't edit room topics, pinned messages can get a bit weird once you start editing them, and I have a few invites to some pretty nasty looking spam dms that I can't delete, so there are still some rough edges.

On the desktop side, if there were a client as good as SchildiChat maybe that would work, but last time I tried one of the Element desktop clients I wasn't even able to log in (it either crashed or hung, can't remember), and most of the time I'm fine with Fractal anyway. Fractal is actually a very nice client for what it does, it just has a limited feature set: missing spaces, copy/paste doesn't quite work like you'd expect, no search (I'm not sure there are any clients with fully functioning search for encrypted rooms), and my memory is that heic image previews weren't supported. I can fall back to nheko for some of the other things when needed.

As far as I know, there aren't any clients that support the new element call unless you enable the labs feature in element x.

All that said, I can't overstate how much I appreciate all the work the matrix devs do, and it is still fine for my daily use. Even if I sometimes disagree with Arathorn's conclusions about how ready matrix is, I have to appreciate the optimism and I imagine it is part of why he's been able to continue through all the negativity :) and it's not entirely wrong to say that matrix beats the competition - I'd say it easily beats teams and imessage (teams does not deserve to be mentioned in the same sentence as anything else), and it is mostly comparable with messenger and whatsapp. Slack probably has it slightly beat, and discord is leagues ahead of everything else.

I think an early major mistake was that Matrix spent all this time and energy designing a general synchronization protocol, but not doing it in terms of native encryption and cryptographic identities. This was post-Snowden, and it was glaring at the time.

Bolting on encryption after the fact then sucked so much energy out of the ecosystem for clients. This one doesn't implement encryption, that one does but it has bugs/warts/etc, this other one does if you pull this year-old experimental branch and build it yourself. The web (-technologies) client became the de facto one because it "worked", despite being bloated and laggy - reliable tools don't even have the code to show these spinning delay circles that have become synonymous with the web ecosystem.

I don't want to be entirely negative because I do see it as the least-worst messaging option available. I use it for communicating with a bunch of friends and things do seem to be getting better, and I look forward to when I can actually switch to its window to type a message and not wait around for redraw / garbage collection / reloading messages from server /etc. That might be on me for not having surveyed native clients recently or tried Element X on my desktop, but that's exactly the negative momentum I'm lamenting above.

That's true, but in another sense I think it's just that trying to "do it all" encryption-wise is significantly harder than some people realize. Having encryption that's really "safe" raises barriers to casual use that most casual users aren't really willing to accept.

Like suppose I use Matrix only on my phone, so I just have the one device. Then I lose my phone and have to get a new one. How do I regain access to my account, including all of my old messages? Or suppose I (still using Matrix only on my phone) decide I need to log out because I want to let someone else (a friend, my kid) use my phone for a while and don't want them snooping in my messages. How do I retain access to all the messages I receive while logged out?

I'm not saying these are problems with Matrix; they would be problems with any service that attempts to cover the same bases (in particular, e2ee with forward secrecy and multiple independent devices). The average user's conception of a messaging service is "I can log in with my password and then have total access to all of my messages, past, present and future." There are too many ways to break that assumption if you try to have perfect forward secrecy and all these other desiderata that encryption wonks care about but normal people don't. I think this is one reason there's still a big gap between comments on HN saying "matrix still works fine for me" and the tales of "I tried to get my grandma to use this and it was a disaster". I don't think it makes sense to try to roll Matrix out for general use, or say it's "the best messaging app" until it can smoothly handle all of those use cases.

I'd say that all of those usability problems are made worse by bolting on encryption after the fact. It makes it so there are now two layers of authentication/identity, rather than a single one. Whereas with built in encryption, you can always punt on solving the problems you list and make implementations that are less secure. For example, just store the keys on the server, accessed by the password. That's obviously less secure, but not less secure than no encryption.

Yeah, I don't disagree. Just saying that even if they had put encryption in at the outset, there are still tough issues to solve in terms of integrating that with user expectations.