Matrix is doing pretty well (although I might be biased, being project lead). For instance, the entirety of the German military announced moving over to Matrix using a fork of Element a few days ago: https://sifted.eu/articles/european-armies-matrix/, and there are several other (very) major governments beyond France & Germany who are switching to Matrix in order to get end-to-end-encrypted interoperable comms that they can run themselves. Hopefully we should be able to announce them in the coming months :)
Meanwhile, we're about to finish the main milestones of making Gitter natively speak Matrix (planning to announce on Thursday), and on the FOSS side in general, Mozilla has successfully moved over entirely to Matrix for community facing work (https://matrix.org/blog/2020/03/03/moznet-irc-is-dead-long-l...), and a few other similar sized open source projects are in the process of finalising doing the same manoeuvre; watch which this space for updates. https://element.debian.social/ looks to be a thing, for instance :)
In terms of naming, it's really not hard: Matrix is the protocol; the core team set up a company called New Vector who made a flagship Matrix client called Riot, but we then renamed both the company and the app to be Element in order to simplify things. Hopefully most people have forgotten the renaming and moved on by now.
In terms of what's next for Matrix:
* Loads of effort making Element more mainstream-friendly; going through improving the UX and making it consistent cross-platform and predictable for new users. The approach we're taking is to film new users using the service, and literally tight-loop fixing the thinkos that they trip over until they stop tripping over. It sounds obvious, and we should have done it years ago, but it's starting to make a big difference. The first wave of changes ship in the next 2 weeks.
* Social login - implementing OIDC Connect to simplify onboarding if you're willing to hand over your identity to an existing identity provider (Github/Gitlab/Apple/Twitter/Google/FB)
* We're in the middle of implementing Spaces - super fun feature to define arbitrary hierarchies of rooms; a bit like discord servers, or slack workspaces, or possibly a usenet hierarchy or IMAP folder tree depending on how you squint. We're hoping to get the first cut out by the end of the year, which is super fun.
* We're also working hard on freeform threading (you too could implement HN/Reddit/Email/NNTP/Twitter on Matrix if you want!). https://github.com/matrix-org/cerulean has some details for the intrepid.
* Peer-to-Peer Matrix is going great guns; we've just finished the first cut of a new P2P overlay network called Pinecone (based on Yggdrasil, but using source routing).
* We've almost finished a wave of work to make 1:1 VoIP not suck; fixes are already shipping in Element on all platforms, but we're almost at the point where VoIP is robust and reliable rather than a quick proof-of-concept which we'd not had a chance to ever really polish.
* Finally, lots of work queued up to make end-to-end encryption more usable. Particularly, chasing bugs where encryption fails (we just fixed a major one in iOS for instance, which shipped a few days ago, thanks to the Push service extension sometimes racing with the main Element process desyncing). We're also looking at simplifying the E2EE key recovery process and just switching to using the same password to both login and decrypt your messages, rather than separate login password & security passphrases as we have today.
I could go on, but TL;DR: I think "the only one with an open protocol seems to be Matrix, but it hasn't proven to be to their advantage, apparently." is bogus. We'd just be another random open source chat webapp if it wasn't for Matrix; instead it's a global open network with 25+ million users and about 60,000 servers. So not yet as big as Email or the Web, but bigger than (say) bitcoin, and continuing to grow exponentially.
Just out of curiosity: the press releases I've seen about these big government players switching to Matrix generally mention that they're using a customized client of some sort. That's understandable, but what I want to know is: Have they also customized the underlying protocol by which their homeservers communicate?
The reason I ask is that it seems to me (as an entirely uneducated outsider) that if they're using the Matrix protocol as-is, then they have incentive to support and assist with the further development of the Matrix protocol, which is great. But if they've already started customizing it, then I would worry that they eventually will decide they don't need Element or the main Matrix protocol and will just go off and do their own thing.
Nope, none of them have customised the protocol (at least of the ones we’re aware of and work with). They realise the whole benefit of Matrix is to speak Matrix rather than some mutant thing.
> "In terms of naming, it's really not hard: Matrix is the protocol..."
First time I read that it was clarifying. Second time I read it it reminded me - naming is important. Always think about customers and don't dismiss things that confuse customers / end users. I 100% want to see y'all succeed. But this customer friction area and similar ones are not to be overlooked. If I were you, I'd focus 100% on the best way to surface Matrix's value to others. That's all we're looking for.
is awesome. A CLI client that is reliable and cutting edge would be awesome (looks like some contenders - I'd love to see one in Clojure or Rust but that's just me..). Bridges with other providers like Slack etc. would be a good way to make in-roads into the Slack user base - would get developers using it for work, and then able to switch over, etc.
Most users of Slack, etc., are getting a little tired of it. We'd love something new - and something we could customize more! Awesome. But you have to position yourself as: (a) a strong, stable, growing alternative with a strong customer focus and appreciation (knowing that if you want to be Linux v. Microsoft and not GNOME/KDE v. macOS/Windows - i.e., succeed in the Enterprise outside of Europe, you have to be a new kind of customer obsessed than what has come before) (b) a low risk alternative to start experimenting with.
Biggest focus should be developers creating enterprise apps (like the German army). That's the sweet spot it seems like to get exponential traction. If you can tap into engineers who are using Slack/Teams/etc. and get them to build more clients, etc... it could be really good.
I.e., implement the difficult parts of Slack really, really well (again be Linux OS v. Microsoft OS - and let hypervisors and whatever else innovate on top of your platform -- and not try to boil the ocean with UI features, etc. - which I feel like is just difficult to compete with - there's a mountain of engineers in Cupertino who are going to out-innovate on a stronger platform with a lot more advantages.. but something like the Linux kernel.. is key).
> A CLI client that is reliable and cutting edge would be awesome (looks like some contenders - I'd love to see one in Clojure or Rust but that's just me..)
Weechat is really impressively good on Matrix, and the plugin is in the final stages of being rewritten in Rust. It has full E2EE support, and the developer is funded to work on it full time.
> I.e., implement the difficult parts of Slack really, really well
Yup. The balance between features and core stability is Hard, but hopefully we're getting the balance right (at last).
Hydrogen (hydrogen.element.io) is an example of a ludicrously lightweight, stable, but not-yet-featureful Matrix client fwiw, as an example of what the balance looks like for "implementing the difficult bits really really well".
Meanwhile, we're about to finish the main milestones of making Gitter natively speak Matrix (planning to announce on Thursday), and on the FOSS side in general, Mozilla has successfully moved over entirely to Matrix for community facing work (https://matrix.org/blog/2020/03/03/moznet-irc-is-dead-long-l...), and a few other similar sized open source projects are in the process of finalising doing the same manoeuvre; watch which this space for updates. https://element.debian.social/ looks to be a thing, for instance :)
In terms of naming, it's really not hard: Matrix is the protocol; the core team set up a company called New Vector who made a flagship Matrix client called Riot, but we then renamed both the company and the app to be Element in order to simplify things. Hopefully most people have forgotten the renaming and moved on by now.
In terms of what's next for Matrix:
* Loads of effort making Element more mainstream-friendly; going through improving the UX and making it consistent cross-platform and predictable for new users. The approach we're taking is to film new users using the service, and literally tight-loop fixing the thinkos that they trip over until they stop tripping over. It sounds obvious, and we should have done it years ago, but it's starting to make a big difference. The first wave of changes ship in the next 2 weeks.
* Social login - implementing OIDC Connect to simplify onboarding if you're willing to hand over your identity to an existing identity provider (Github/Gitlab/Apple/Twitter/Google/FB)
* We're in the middle of implementing Spaces - super fun feature to define arbitrary hierarchies of rooms; a bit like discord servers, or slack workspaces, or possibly a usenet hierarchy or IMAP folder tree depending on how you squint. We're hoping to get the first cut out by the end of the year, which is super fun.
* We're also working hard on freeform threading (you too could implement HN/Reddit/Email/NNTP/Twitter on Matrix if you want!). https://github.com/matrix-org/cerulean has some details for the intrepid.
* Peer-to-Peer Matrix is going great guns; we've just finished the first cut of a new P2P overlay network called Pinecone (based on Yggdrasil, but using source routing).
* We've almost finished a wave of work to make 1:1 VoIP not suck; fixes are already shipping in Element on all platforms, but we're almost at the point where VoIP is robust and reliable rather than a quick proof-of-concept which we'd not had a chance to ever really polish.
* Finally, lots of work queued up to make end-to-end encryption more usable. Particularly, chasing bugs where encryption fails (we just fixed a major one in iOS for instance, which shipped a few days ago, thanks to the Push service extension sometimes racing with the main Element process desyncing). We're also looking at simplifying the E2EE key recovery process and just switching to using the same password to both login and decrypt your messages, rather than separate login password & security passphrases as we have today.
I could go on, but TL;DR: I think "the only one with an open protocol seems to be Matrix, but it hasn't proven to be to their advantage, apparently." is bogus. We'd just be another random open source chat webapp if it wasn't for Matrix; instead it's a global open network with 25+ million users and about 60,000 servers. So not yet as big as Email or the Web, but bigger than (say) bitcoin, and continuing to grow exponentially.