XMPP Council - 2017-10-18

is it that day of the week again?

Indeed it is.

Yep, and I’m not ill anymore!

1) Roll call

o/

daniel, SamWhited, Dave

Sorry, on 4G with conversations and it was trying to join the MUC with an disconnected account.

Here

oh. yes i'm here

yay...everybody here

forgot that it's wednesday again

2) Minute taker

any volunteer?

I can't I'm afraid; on a tablet and it's not very easy to multitask.

daniel: it's Wednesday every week at this time

shall I?

Ge0rG, thanks

3) Vote on accepting https://github.com/xsf/xeps/pull/528 " XEP-0071: make security considations much clearer #528 "

+1, there are other things I want to add next but it already is a net improvement.

+1

+1

+1

+1

thanks

4) Vote on accepting ProtoXEP: Body Markup Hints https://xmpp.org/extensions/inbox/bmh.html

I'll vote on list

I'd like to see how the discussion goes on this one.

I'll interpret that as 'on list' :)

I’ve already read it yesterday evening, and I’m very much -1 on it due to the concept of waiving any format support, forcing implementations to support multiple formats and making it impossible for a message to carry more than one (think MUC for example).

You can interpret it as not voting yet, indeed.

i'm not a huge fan of the XEP personally but technically it looks fine. so +1 from me

I'm very much waivering somewhere between Link Mauve and daniel's positions, so on list for me (though it might not matter if Link Mauve is -1)

alright

(My -1 might change, but not with the XEP as is.)

Link Mauve: is "very much -1" a -1 vote?

Yes. :)

I'd appreciate a debate on list about this, mind - I'd like to explore how we might change your mind.

5) Vote on accepting ProtoXEP: Atomically Compare-And-Publish PubSub Items https://xmpp.org/extensions/inbox/cap.html

I'm on list on this too

I've not had a chance to review this properly yet, so I'll vote on list later.

dwd, sure, but it’s already been mostly summarised in the long XHTML-IM thread, which I also read last night.

Bad luck. :-)

Tobias, on list too.

On list

+1

There are a few things I don't love about this one, but I think having the operation is useful and it's a good start.

ok

6) Vote on accepting ProtoXEP: Jinge Encrypted Transports - OMEMO https://xmpp.org/extensions/inbox/jet-omemo.html

also on list here, as I haven't read it yet

On list too, for the same reason.

Jinge encrypted?

on list

dwd, just a typo here. :)

on list

I think I'm +1 on this, given Paul's implemented it so the real errors are probably in spec not fundamental.

ok

The rest I think is pending external events or further discussion. Will ping respective people afterwards.

7) Date of next

same time next week

What about obsoleting XHTML-IM? I know that's the controversial one, but it seems like the discussion has moved on to "what to do next"

So I don't see any reason to hold off obsoleting it and indicating that we want to do formatting-nextgen

I mentioned one

but you might disregard it of course

i haven't had time to read through the thread

SamWhited, we just started the discussion a week ago, i suggest waiting for next week to vote on obsoleting it

SamWhited, jonasw’s email was very detailed about why we shouldn’t do that without a path forward.

I am still not keen on obsoleting XHTML-IM before we have an actual alternative ready. I don’t think that this will achieve anything good. Instead, I think that one of two things will happen: (a) Clients continue to implement XHTML-IM because it is the only actual way to convey markup right now (this is what I’ll do until there’s a replacement). (b) The ecosystem will fracture in islands of different, underspecified, plain-text markups put in <body/>. I don’t think either is particularly good. I also wonder what it would look like to have the only markup protocol with actual deployment being obsoleted :-) (*hint towards the general direction of the Experimental vs. Draft discussion*).

like daniel, i haven't read all the mails in that thread yet

Yeah, I think people have raised objections largely predicated on "what else would we do", though. So I don't think there's community consensus just yet; this despite my feeling that we should deprecate it.

Let's discuss next week then

Ack.

so time for next, as usual

8) AOB

Wfm.

AOB - I noticed the email from Klaus; we should probably consider that *before* someone demonstrates the problem...

Perhaps simply [re]confirming votes on list in this room would be enough?

While making sure we are using our usual JID?

Sounds like premature optimization. I can't imagine that it ever becomes a problem, and even if it did it would be pretty easy for someone to see it and say "wait, I didn't send that"

Reminder for daniel to to vote in the "[Council] 2017-09-27 XSF Council Minutes" thread on accepting "XEP-0060: Add pubsub#multi-items in Publish-Subscribe features #500"

SamWhited, that was the answer I was going to make after this meeting, that we should be aware of the emails “we” send.

+1 on what SamWhited said

Basically monitor the mailing list.

Oh, and only us can post to council@, correct?

That may well be sufficient, too.

SamWhited, given that nobody reads all emails on list, I find that a not-so-good solution.

GPG would be nice

Link Mauve: Yes, and thankfully email can't be forged ;)

Yes, but that means anyone using your email address.

jonasw: Council do, though, that's kinda the point.

Kev, okay, if you say so :)

Kev, do you mean the ML software isn’t checking my server is actually who it pretends to be? :(

Good luck with finding a standard for that.

jonasw, I make a point in reading everything, even if I don’t always answer.

dmarc will solve everything.

Link Mauve, how would it? :)

SPF would be an option, sure...

But yes, I think that's the worst bit of being on Council, the responsibility to process every single blasted mail.

email is a mess.

Kev, didn’t know that it was a requirement.

another point for AOB, Ge0rG's very old PR https://github.com/xsf/xeps/pull/434, it still says "Needs list discussion". Ge0rG, is that still correct?

Tobias: Hints was -1ed, so that PR is in limbo

Limbo?

Ge0rG, would it get unblocked by your suggestion of better semantic hints?

Tobias: I'd like to hear if the council is +1 on the proposed wording change, to make the rules stricter

I'm still -1 on enforcing the rules, FWIW.

Kev: what would make you change your mind?

Ge0rG: Having the big picture clear and agreed first.

I think by the way that we urgently and before advancing MAM&CSI need to have a discussion about the point georg brought up about message routing.

Kev: so I should continue with my TLDR posts?

I want to take part into that big whiteboard routing party, where do I sign?

TLDR is always nice

Ge0rG: Absolutely. And I should read the one you sent the other day.

Kev: that's not very motivating.

I agree, I like the idea of clarifying the rules, but I think it needs a bit whiteboarding party first. I don't have confidence that this is necessarily the correct set of rules.

I promise I'm going to, if that helps. But, busy week.

*big

I’m all in for whiteboarding, if I can attend somehow :)

ok..any further AOB topics?

doesn't look like it

thanks everybody

*whew* busy meeting; thanks all!

thanks Ge0rG for taking notes

Thanks all. :)

please send them to standards@ and council@

Link Mauve, "waiving any format support"?

btw: people interesting in serving on council, please add at https://wiki.xmpp.org/web/Board_and_Council_Elections_2017

is nobody gonna re-apply?

Flow, just like 0380, you are making an incomplete list of supported formats, clients are free to pick any of them (as long as there is disco, otherwise it’s outright impossible), and there is nothing required to be supported.

jonasw, probably people are too busy and haven't added them yet

disco doesn’t work, I should’ve added that

(to my email on that protoxep)

<body-markup-hint language="text/html"/> is totally fine, as an example.

oh the pain

Link Mauve, I don't think the situation is comperable to xep380

that sounds so much like things libpurple would do.

jonasw, Tobias, it isn’t the day before the deadline, nobody has even started working on their application. :p

So, the situation I'm is that I've data which is formated using CommonMark. And I want to send that data to an XMPP client. I don't want to make the effor to write a CommonMark to XHTML-IM-NEXT convert, so I just shove it into the <body/>

I find that a super bad idea

you shouldn’t be doing that in the first place

And all I want to achieve with BMH is to tell the receiving entity: Hey, there is CommonMark in the <body/>

Flow, not making the effort of using a library to generate a common format that everyone understands is a very bad practice, imo.

jonasw, but I will do that, and others will also

either you care about interop (and use a separat element + plaintext variant in <body/>), or you don’t

tacking a "greenwashing" "oh and if you happen to support my unspecified format, this is it" tag doesn’t make things a whole lot better

If you consider that CommonMark to be the plain-text alternative of the XHTML-IM version, you can put it in the body already, but please don’t assume the recipient will also need to support CommonMark.

I think <body/> is the highest level of interop

putting markup in <body/> isn’t

So I do care very much

Yes, I agree on that.

jonasw, exactly.

Link Mauve, I don't assume that

but I give him a hint that it's CommonMark

putting marked-up things in body could easily break accessibility tools

jonasw, …

Flow, if my client doesn’t support CommonMark but reST, will you convert that for me?

Flow, …?

FWIW, I’ve heard "there are no accessible clients for jabber" more than once as a reason not to use it.

Link Mauve, If the data I have is already formated in CommonMark, possibly not

while many of the hip messengers in fact are quite accessible.

If you are talking to me and my friend there whose client only supports Creole, in a MUC, are you going to provide us two different messages?

I feel like the XEP would had more acceptance if there was no disco part

the disco part is irrelevant, because it doesn’t work

Link Mauve, that's another aspect

it doesn’t work in MUC, it doesn’t work in MIX, it doesn’t work in modern one-on-one chats due to carbons and MAM

Flow, all that just because you couldn’t find a library to generate XHTML-IM from your CommonMark?

(also, shouldn’t we move this discussion to xsf@?)

Ge0rG: Because I'd hate to demotivate you.

(I’m some 2000+ lines up in that buffer, I have a lot of backlog to read before that discussion then. :x)

(Damn being ill!)

:(

Sorry, I just got a bunch of work calls. Now working up the AOBs.

dwd: what's the context of "dwd> AOB - I noticed the email from Klaus; we should probably consider that *before* someone demonstrates the problem..."?

Ge0rG: Voting on list without sender verification.

can I write that into the Minutes?

It was an AOB, yes.

I am writing up a reply to the thing Link Mauve suggested hadn't been addressed (jonasw's email), FYI. It will be on list shortly. I think I've addressed all these points individually, but maybe it wasn't clear.

I'm not sure if my mail to council@ went through, or if I need to fake somebody's email address.