[Standards] Review of XEP-0369: Mediated Information Exchange (MIX) v0.9.5

Sam Whited sam at samwhited.com
Sat Mar 24 15:42:35 UTC 2018

On Sat, Mar 24, 2018, at 00:31, Evgeny Khramtsov wrote:
> I think the core of the XEP should not rely on ad-hoc services: the
> whole protocol should re-use existing pubsub spec. For relaying
> messages this would be enough. Other stuff like presence management,
> jid proxy, access rules and other nerdy features should go into an
> ad-hoc service and should be described in a separate document.

Without looking at the specific features mentioned, I agree with the sentiment of this.  We had a limited set of use cases when the first draft of MIX was written, but it appears that we've gone far beyond that now. There is a lot of unnecessary cruft that's not necessary for group chat in MIX; it's already hard enough to build since it requires pubsub (which is itself rather bloated), let's not add more on top of that unless it's absolutely necessary.

That being said, I'm not sure what's worse at this point: how large the spec has gotten, or yet another large change.

> Fri, 23 Mar 2018 20:57:41 +0000
> Kevin Smith <kevin.smith at isode.com> wrote:
> > MIX is a very simple core concept

That has rarely if ever counted for anything. The biggest problem XMPP has (and I think I do seriously mean "biggest", not just in a hyperbolic way) is that it takes simple concepts and invariably overcomplicates them.

> > and I’m not entirely sure how we’ve got to the current situation
> > where it’s viewed as complex
> The XEP is 91 pages long. Even reading it will take a day.

Though length isn't necessarily an indicator of complexity, I agree that this is a big part of the problem. It doesn't matter if it's actually simple if no one wants to read it. If you rewrite War and Peace a lot of people aren't going to read it no matter how good it is. To a certain extent I expect a group chat XEP to be longer than others: it's a big topic. However, as mentioned before MIX just seems unnecessarily long.


More information about the Standards mailing list