This is a nice perspective, especially with the brief comments that follow the main post. A part of me wants to disagree with the sentiment the author expresses in the last comment, but more and more I'm finding it hard to deny:
> the problem you're [i.e., the Matrix project lead is] describing isn't actually the problem that Matrix users have, and the solution you're building isn't actually the solution that Matrix users want
This can be seen in various slogans and visions from the heady earlier days of Matrix. It was talked about as a one-stop solution for real-time communication, etc. They even had a prototype of some kind of "metaverse" thing.
But the thing people are looking to Matrix for is chat. That's it. Not an "eventually consistent real-time JSON database" or anything like that. Just chat. And, like the author of this post says, choosing a DAG as the bedrock data structure may make it hard to provide some of the bedrock functionality that people are looking for in chat.
I will say that some of the other issues mentioned do seem at least partially solvable within the framework that Matrix has set up. Like, servers could get faster, moderation tools could get smoother, etc. It's still true that right now those problems haven't been addressed, but it's possible that, from the user perspective, things could improve a good deal. The biggest question mark for me is how moderation can really work in a decentralized system like this, because it's not clear to me that we really have a workable model for that even conceptually (let alone that Matrix is it).
17
2