XMPP Council - 2024-08-13

Meeting? Or have I been timezoned again?

Oh. Yes. Sorry. I was distracted

feels a bit weird to start 20 min late. and apparently nobody aside from moparisthebest noticed. i'll make sure to prepare a proper agenda next week. however can everyone (async if you wish) let me know if you are happy with the current version of https://github.com/xsf/xeps/pull/1365/ (That’s a MUST + note on the destroy element) ✏

(indicate via a +1)

+1

if that’s the case I can merge Guus' MUC PRs over the course of the week and we'll continue with other stuff next week

daniel, as I'm still working on the project that caused me to trigger all these changes, more could follow (but I'm not sure). I have no issue with the changes going in a couple of different revisions, but I'd also understand you wanting to push them all out in one go. If that's the case, we might want to coordinate a bit.

"CAN NOT" is not defined in RFC 2119, maybe it should be lowercase? Or SHOULD NOT ?

+0

MUST NOT is the right formulation there, isn't it?

Server MUST send and client MUST NOT expect it seems odd...

MUST NOT rely on it, though.

I opted for CAN NOT (which indeed should be lowercase) instead of MUST NOT intentionally, as I thought it weird for there being no alternative available to detect the event (that I can think of).

What about replacing `CAN NOT depend` with `cannot reliably depend`?

That's fine with me too

It being normative is needed for interop, though.

"As this requirement was added in a later version of the specification, entities MUST NOT rely on a MUC service including this" or thereabouts is needed if you don't want it to be a breaking change. ✏

Is the event distinguishable at all as a 'destroy' event, if that element isn't included?

NAFAICS.

then interop is broken anyway, right?

No, in current 45, you can't tell if it's a destroy or not, and that is known.

as an aside, I still believe that 45 always mandated this, given this text in 10.1: > The service MUST then destroy the room, sending a presence stanza of type "unavailable" from the room to the owner including a <destroy/> element and reason (if provided) as defined in the Destroying a Room section of this document.

Isn't your text adding that requirement to non-owner presence, or do I misremember the PR?

In any case, I hear not work calling.

No, you're right - I am assuming similarity.

Although I do find the "MUST NOT rely" wording a bit odd, I've got no issue with using that instead of what I proposed a few lines ago

whatever council prefers works for me.