XMPP Council - 2019-05-01

Hi, I again won’t be able to attend tomorrow’s meeting, this time due to the International Workers’ Day protests, I’ll attempt to send my pending votes by email before leaving tomorrow morning but I can’t guarantee it.

I’m sorry about that. :(

While my simulacrum is present in this room, it doesn't mean I can attend. Please start without me if I don't give a sign.

Do we have agendums for today?

not as far as I know

Kev, Sorry, I've been finalising a new position and got a little distracted.

I think there was something leftover from last meeting though

Re-voting on #758

There was a pending 0184 AOB from last meeting.

1) Who's Here?

✋

Link Mauve, I see your apologies.

Ge0rG, You're not meant to be - didn't you say you were skipping this one?

2) Agenda Bashing

dwd: sorry, I'm calm now.

Apologies for the missing agenda, as my current unemployed status might suggest, things have been impressively complicated for me. Should be settling down.

apology accepted and all the best for your employment future :)

So, we need #758 and '184. Anything else?

I don’t think so

OK.

3) Items for a vote:

a) #758

https://github.com/xsf/xeps/pull/758

+1

XEP-0060: Expose pubsub#access_model and pubsub#publish_model

+1

IIRC I was +0 on this one.

Ge0rG, Are you still?

Yes

OK.

Kev, ?

This is the one we discussed last week and decided was ok and I was the only previous blocker, which I was now removing?

+1 if so.

yes

Kev, This is indeed.

OK. Link Mauve on list.

'184 is not, as I recall, a voting matter?

No, a discussion one.

b) Discussion: XEP-0184

Ge0rG, Take it away.

The small fig leaf box in 0184 is the attempt of introducing the Right Behavior without normative language. Do we want normative language in there? Shall we introduce it with a bump?

*IF* we are going to bump '184, I want multiple acks per message

https://mail.jabber.org/pipermail/standards/2019-April/036090.html

jonas’: that's an excellent point.

If we were going to bump 184, doing so in a manner that allowed MAM/archive smarts, as well as MUCciness would probably be sensible.

and I think a bump might be worth all of the benefits combined

Kev: what kind of smarts, explicitly?

Kev, I *think* I'd rather do those upgrades in XEP-0333, which is largely similar but has more capability already.

The metadata consolidation dance we talked about at the Summit (or were you not there for that bit?)>

I think 333 needs careful burning to the ground and recreating. But I agree with the sentiment of doing things in a general way.

I don’t think XEP-0333 is a good idea at this point, but that’s a different discussion.

I think that adding the strawman text from my post doesn't _require_ a bump, and I'm not going down the "if we bump anyway" route any time again after Carbons.

Ge0rG, +1 to that, indeed.

Which of the three possible things you could be +1ing are you +1ing?

Ge0rG's text, Ge0rG's assertion that it doesn't need a bump, or Ge0rG not wanting to 'if we bump anyway'?

Kev, The argument against "If we're going to ${MINOR CHANGE}, we may as well bump and do "{COMPLETE REWRITE}".

If we can make things better without a breaking change, I'm in favour of doing so.

I agree

I would be opposed to a breaking change that didn't solve the bigger issues, I think.

It's up to The Elders to decide whether my text is a Breaking Change

Also what about headline and error?

error should be MUST NOT

I think I can see my way to this not being a breaking change. Also, error makes sense as SHOULD NOT rather than MUST NOT.

because it is in fact possible that a request is in there, and you don’t want to reply to that (because don’t auto-reply to errors)

Sometimes you might want to know that your error was received :)

hm

Fight it out!

I see the danger for weird loops there

TBH, I'm not hugely against MUST NOT.

I'm with Kev here, but I see the potential for loops as well

but I’m fine with SHOULD NOT

I think the last line about MUST ... different type might benefit from a "(as previous versions of this document didn't fully specify the required types)".

I don’t have a strong opinion about "headline"

Otherwise one might read it and wonder why we bother specifying the types if we then don't care about them.

Kev: good point.

I'd also probably SHOULD instead of MUSTing it.

Kev: why?

Just feels right, I don't think I have a solid argument.

That would indeed break compatibility

Not a hill for me to die on, just curious.

I can live with MUST there.

While reserving the right to be smarter next week than I am this week, and change my mind ;)

Any other suggestions? I would prepare a PR for a future council meeting then?

Ge0rG, do that

Thank you very much!

Sounds good.

I can't promise I won't have other minor comments in light of a PR, but that sounds like a good direction, thanks.

4) Are we all up to date with voting?

I believe I am, at least.

We are missing link's votes on a bunch of changes where I pretend they don't break anything

I think Ge0rG has a vote on #736 that expires today.

according to tedds summary I am

And I'm still waiting for link to provide implementation guidance on server side LMC

jonas’, Really? I don't see - am I blind?

I am confused

you mean you don’t see the email? Subject: [Standards] Council Voting Summary 2019-04-30

came in at 2019-05-01 00:22:00 UTC

dwd: is 736 the MUC LMC enforcement?

jonas’, No, I see the email, I don't see what vote you're missing.

Ge0rG, Yes. Expiring today, so if you need feedback from Link, you should veto.

dwd, none, I think

Ge0rG, Well, that said, everyone else vetod, so..

that’s why I said "I am [up to date with voting]"

Okay, I'm -1. Not that it would change anything

jonas’, Never mind - I'm going mad.

Reason: unclear performance impact on server implementation

OK.

5) AOB

I am undergoing an affiliation change, FYI.

/affiliate dwd owner

outcast->member

That sounds like no AOB.

Indeed.

6) Next Meeting

+1 wfm

Next week is achingly clear for me.

+1W

SBTSBC should WFM

Excellent.

In that case:

6) Ite, Meeting Est.

And thanks all.

Thanks, dwd

Thanks all.

thanks :)