Re: [Tools-discuss] Trial chat services: matrix and zulip

[Ted Lemon <[e-mail]>]

On Oct 6, 2020, at 12:05 PM, Michael Richardson <[e-mail]> wrote:
> To me, it feels that XMPP is being called a failure by the entities that want
> to sell us something else.
> For me, it's been a great success.

There’s no conspiracy here. XMPP has had a really high barrier to entry for use in the IETF for a long time. I’ve been asking for the IETF to run an xmpp server for a long time, and the answer has always been “no.” We’ve always been pointed to XMPP servers that have no SLA, and are operated for free on a “good luck, don’t bug me” basis. This basically means that anyone who doesn’t have a couple of hours to spend getting XMPP working is going to have trouble.

On top of that, even with a dependable server, which we don’t currently have, the clients have been languishing for years and are not being maintained. If I want to get a reliable xmpp client going on my Mac, I’m going to have to download and compile code, and I don’t actually know which code is the best bet, so chances are I’m going to have to download and compile multiple bits of code before I find one that works.

This is a really heavy lift for something I’m going to use for maybe ten hours per IETF, three IETFs per year.

If we think XMPP is really the best thing, we should spend money on it and make it actually work well, rather than just telling people to use it, and good luck.

I tend to agree that we want something that’s an open standard; Matrix seems to be a good choice in that regard, but I admit I haven’t kicked the tires sufficiently.