Features like these make me wonder what the goal and direction of Matrix really is. I can see how this is a nifty demo of how the Matrix protocol can support many use cases, but the base use case (open communications) isn't nearly complete yet.
Element is the only application supporting all Matrix features, but it's look and feel are off, it's speed leave things to be desired and I'm still getting issues with fetching the right keys for a chatroom on some of my devices. The file picker on mobile is wonky, some very basic features (like adding a caption to an image you're sending) just don't work. I'm definitely not convincing anyone to consider this over the existing, isolated, closed-source messaging applications out there.
Then there's the fact that there's very little chat clients that support most of the protocol, especially with encryption. To combat this, you can run pantalaimon if you can ever get it to work (its dependency on things like dbus makes it very hard to run well on a VM or in a container) to add encryption to more well-integrated chat clients like Fractal.
I love the premise and the idea behind Matrix, but I do fear that the focus on expanding and experimenting will eventually prevent it from reaching the critical mass necessary to ever be used outside of the enthusiast circle. All the (optional?) additions to the protocol also make me fear that Matrix will end up dead like XMPP, with different features implemented across different applications all in a way that things just don't work well together.
Why put all of this work into running a daemon 24/7 on your phone that will end up sucking battery to run an app with bad usability in peer to peer mode? What's the end goal? I just want to rid myself of Whatsapp and Telegram, is that even a valid use case for the Matrix system anymore? Should I start looking in other directions?
The reason we occasionally do experiments on Matrix (VR, 3D video calling, low-bandwidth, P2P and Cerulean) is to try to remind folks of the broader use cases of Matrix beyond chat and to encourage contributions and experimentation from the wider community, so these ideas can evolve outside the core team. In the case of Cerulean, it's also as a test jig to test out threading in Matrix for use in Element and other clients.
Meanwhile, just because one person (kegsay) spent a few weeks putting together Cerulean doesn't mean that the other thousands of people in the Matrix ecosystem were idle. For instance, Fractal merged matrix-rust-sdk support 2 weeks ago, which implements full E2EE (https://gitlab.gnome.org/GNOME/fractal/-/merge_requests/626), so your complaints about pantalaimon should shortly be irrelevant. (Plus it seems pretty miserable to complain about pantalaimon's dependencies; the fact that the project exists at all as an E2EE stopgap is unusual - XMPP has no equivalent for instance, as far as I know). In terms of other clients supporting E2EE: weechat-matrix, Mirage, FluffyChat, Seaglass, Nheko, Gomuks, Daydream and more have support. Complaining that there are clients which don't have E2EE is like trying to denounce the Web because there are browsers out there which don't implement HTTPS.
Ironically, Cerulean is the first Matrix client to implement captioned images (after a fashion), so your captioned images on Android might benefit directly from Cerulean's work there...
> Why put all of this work into running a daemon 24/7 on your phone that will end up sucking battery to run an app with bad usability in peer to peer mode? What's the end goal? I just want to rid myself of Whatsapp and Telegram, is that even a valid use case for the Matrix system anymore? Should I start looking in other directions?
XMPP isn't dead at all. OMEMO offers all the benefits of Signal's double-ratchet encryption and it's supported by at least one good client on every platform. Both ejabberd and Prosody are actively maintained servers, and there are still over a hundred XMPP hosts with open public registration. [And countless other hosts that aren't public.]
And it's way way way faster than Matrix. The servers are lightweight and don't require tens of GB of RAM, and the modern clients (like Dino and Conversations.im) are quick and snappy.
Element is the only application supporting all Matrix features, but it's look and feel are off, it's speed leave things to be desired and I'm still getting issues with fetching the right keys for a chatroom on some of my devices. The file picker on mobile is wonky, some very basic features (like adding a caption to an image you're sending) just don't work. I'm definitely not convincing anyone to consider this over the existing, isolated, closed-source messaging applications out there.
Then there's the fact that there's very little chat clients that support most of the protocol, especially with encryption. To combat this, you can run pantalaimon if you can ever get it to work (its dependency on things like dbus makes it very hard to run well on a VM or in a container) to add encryption to more well-integrated chat clients like Fractal.
I love the premise and the idea behind Matrix, but I do fear that the focus on expanding and experimenting will eventually prevent it from reaching the critical mass necessary to ever be used outside of the enthusiast circle. All the (optional?) additions to the protocol also make me fear that Matrix will end up dead like XMPP, with different features implemented across different applications all in a way that things just don't work well together.
Why put all of this work into running a daemon 24/7 on your phone that will end up sucking battery to run an app with bad usability in peer to peer mode? What's the end goal? I just want to rid myself of Whatsapp and Telegram, is that even a valid use case for the Matrix system anymore? Should I start looking in other directions?