I'm never going back to Matrix
shkspr.mobi
shkspr.mobi
Here's how I made it work:
- No federation. You click an invite link I send you and make an account on my server. Therefore, no spam.
- Encryption disabled in the main group chat room. The encryption experience was very poor in Riot and remains inexplicably baffling to non-technical users in Element. Supposedly Element X fixes these problems, but it doesn't support the type of SSO my server uses yet. I decided to just turn it off. DMs are still encrypted.
- Use Synapse.
- $5 tier digital ocean droplet. I plan to move it to my homelab soon.
For the most part, it has been pretty smooth. Dealing with encryption UX issues was the low point.
I wish that there were custom emoji packs though :(
This sounds like how XMPP was about 10 years ago. e.g. the transition to XEP-0280 was rather terrible and anyone complaining about it was told they were using the wrong client; often without any recommendation for which client to switch to.
XMPP and IRC are not it, for me. Neither give me a better experience nor are they easier for non-techies than matrix.
I also empathize with the people behind the project, as monetization is much more difficult for non-scumbag companies, among which I definitely count Discord, Slack and to a lesser degree Telegram.
As a user though, the speed of improvement has been less than satisfying. It has felt like matrix was just shy of fulfilling its promises for years now.
I still enjoy using it though and am hopeful for its future.
What. XMPP is much easier to work with since both servers & clients use an order of magnitude less resources (CPU, RAM, storage, bandwidth). This makes them easier to self-host & also get someone to actually launch & keep an app open if it isn’t spiking. There are handholdingest deployments like the server+client of Snikket. & if you want that web link to send someone that is skeptical of installing yet another chat application, Movim covers that angle with posts, & multi-user, multi-stream audio/voice calls (where you can use the home instance, or self-host it). But also there is clients/services for anything in between—& without a protocol that keeps as much metadata & skyrockets on costs trying to sync the entire history of every chat/attachment for all users (which inevitably leads to all that metadata synced to the mothership, Matrix.org).
Unfortunately, their promises grow at a faster rate than the reality of the protocol and software. The biggest problem they have is that they constantly tout it as this amazing thing that people should start using, when in fact it's got tons of rough edges and it would be a big mistake for most average people to use it. It might eventually get there, but I think it's actually less likely with the kind of self-promotion they do. It leads to too many results like the article linked here, where people go "Wait, you said this would be great but it's actually just kind of barely usable" and they're permanently soured on the concept. It erodes trust in the organization and the product.
It’s impossible to find anything alive with the growing lists of dead old things.
It would be nice for room lists to decay dead channels out of attention from the casual user.
For example, if you’re active in any FOSS channels, you’re likely to receive spam invites to rooms containing illegal content (with disturbing room images and names that appear on the invite). This has been a known issue for years, and a high visibility issue about it (with responses from Matrix’s managing director) from last summer remains open and largely unaddressed.
This issue link is for the Element client, but it contains links to several related proposals for home servers, clients, and the protocol, many of which are still open/completely unresolved. Notably, the MSC related to invite blocking via policy servers or suggestions about ignoring invites via client settings.