XMPP Council - 2017-10-25

Hi daniel, dwd, Link Mauve, SamWhited. It's about time

1) Roll call

Here

Here

Whoops. Here.

2) Minute taker

any volunteers?

jcbrand, maybe you? :)

Tobias: I'd be happy to :)

great

Thanks!

3) Obsoleting XHTML-IM

Where are we on this topic?

I think discussion has stalled. ✏

last message 6 days ago, yes

pretty much cross read everything

It seems like most people on list (with a few exceptions of course) were interested in a replacement, and that a large group (unsure if majority) were interested in deprecating before a replacement is ready given the history of security issues with this spec.

SamWhited, I largely agree; however it doesn't feel like we have consensus across the standards participants, and I'm not sure how we could get it.

I don't think we can or ever will reach a consensus and it's up to the council to make a decision taking the community feeling into account.

Anything this big is impossible to get consensus on, that's why we have a council.

yeah...there are definitly reasonable opinions on wanting to keep it

It did seem to me that most were in favor of a replacement though, and that "go ahead and deprecate" or "push forward on a replacement and wait then deprecate" is the place where there was more contention

Or is that just me?

SamWhited, I don't think that's a bad reading. Just that the minority against deprecation were both vocal and reasonable.

that's also the impression i got

indeed

Cool; obviously I'm biased so I wanted to make sure I didn't misread the temperature of that thread

I think the conversation stalled because some people are watiing to see the alternative.

SamWhited, Don't get me wrong, I'd like to kill XHTML-IM with fire. All the opinion I've got from dedicated web developers is that handling user-entered XHTML is playing with fire.

I thought we were going to see the new spec that wasn't likely to be easy to mis-implement.

I think we should push forward with an alternative and would like to volunteer to write a proposal and start a SIG to investigate how it could be done better.

Kev, I've half a snippets XEP, which might address some of the needs for XHTML.

A proposal for a SIG, I mean.

SamWhited, I don't really want to spin off a whole SIG if we can avoid it.

That also sounds fair, I thought that might be a good way forward but could go either way

I volunteer to email the list asking for interest and requirements gathering then :)

Fwiw I still have that im markdown xep on my todo list

I've been traveling for the last couple of weeks though and didn't get around to

SamWhited, sounds great initating a mail asking for interest/requirements

My plan was to ask for formatting use cases, distill requirements and scope from that, and then see if any of the existing proposals fit or call for a new one.

right, maybe markdown fills those requirements or not, we'll see

Sounds good, I'll prepare that email sometime over the next few days.

SamWhited, so do we want to vote on obsoleting it now or do you withdraw your request to have it obsoleted. Just want to make sure your initial request isn't overlooked.

I would like to request that we obsolete XHTML-IM now. As always, this will not result in everyone not supporting it overnight, it just means we don't recommend new implementations. Given its history, I think it would be irresponsible to continue to recommend it.

arlight..then let's vote on it

4) Vote on Obsoleting XEP-0071: XHTML-IM

+1

+1

I think a greater match for the community would be to have a replacement for the IM case, so I'll vote -1 (noting that this is presumably a simple majority vote and therefore this is not a veto).

Woops, sorry I didn’t see the time.

I’m here now, let me backlog.

i'm also -1 until there is an alternative experimental XEP for the IM case. XHTML-IM is currently the only way to exchange formatted messages between XMPP clients

I'd like to note that I think always requiring a replacement before we can stop recommending something is one of the faults with our process. It's always the same, and is why we have duplicate specs, old things no one implements that are just confusing, and specs with security issues still around.

Also that not recommending it does not mean no one can implement it if they just desperately want compatibility with something else that already has it.

SamWhited, I don't think it *is* part of our process. I just think it matches my sense of the standards list participant's preference better.

Yah, "process" might be the wrong word, "culture" maybe.

I’m -1 on obsoleting, as the XEP fills a very much needed feature and there is no alternative currently.

SamWhited, what do you recommend people wanting formatted messages in their XMPP client and also want to interop with other clients out there right now?

And there probably won't be as long as people can use "but we already have a thing" as an excuse. Obsoleting also provides pressure to come up with a replacement.

I recommend they don't, but if they really need to then nothing stops them from implementing XHTML-IM anyways.

SamWhited, that's true.

Or I recommend they join the discussion about a replacement; obsoleting also provides pressure towards coming up with a good replacement.

SamWhited, pressure isn’t anything we need imo.

If someone wants to take part into that SIG (even informal), they would do so without any kind of pressure from us.

And I will as I mentioned earlier, but that's not the only place where a bit of pressure is helpful.

Actually, I think SamWhited has convinced me to change my vote.

Convincing people to contribute is also useful, taking away "but we already have a spec that's draft" as an argument is also helpful.

On balance, deprecating it will at the very least break the stalemate, so I'll change my vote ot +1 to deprecate.

Sorry for the confusion.

Thanks Dave, I appreciate it. Obviously I feel strongly about it

XEP-0001 doesn't mention whether disapproving XEPs requires a majority vote or all council members being in agreement. But I'd assume it's majority as it's the same for the approval case

> A majority of Council members must vote +1 in order for a XEP to advance.

I assume that's advancing to any state, including obsolete

right

But I'm not sure. Board question?

SamWhited, according to the flow chart it would be an advancement

Tobias, I believe that if Council folk were allowed to veto it would be mentioned.

So with 3 (+1) votes and 2 (-1) votes, we decide to change the state of XEP-0071: XHTML-IM to Obsolete

5) Date of next

Same time next week

?

WFM

I can't do that but happily read the mails

Tobias, Erm, Deprecated, surely, not Obsolete?

so someone else would need to run it

dwd, didn't we have the discussion once, whether to vote twice to Deprecate and then again to Obsolete it or to vote on directly Obsoleting it

or am I misremembering things?

Wfm

I probably screwed up the terminology, I always mix them up. I could see it being either deprecated or obsolete.

Tobias, we did, but dwd said deprecating in this discussion, not obsoleting.

But we are totally ok with advancing twice in a same vote.

I’m still -1 even to deprecating, it’s very much not a sensible direction imo.

Shall we just say deprecated since it should technically happen first and then we can discuss if obsoleting makes more sense next time?

alright..then let's make it Deprecated

i honestly don't care much as I can't see the difference between the two states

6) AOB

yah, I doubt anyone makes much of a distinction

I think the people who voted +1 are happy with obsolete as well

And it wouldn't change the minds of the -1

daniel, yeah...as there's no difference between the two states

Ah right, an AOB, about the pending votes.

Just a reminder.

I could go either way also, so we could just say that dwd gets to decide since he's the only unvoiced opinion?

right, people please vote. Sam and I did today. Someone should takes those votes into trello and move things to editor column if all votes are in

dwd, do you care much whether it's Deprecated or Obsolete?

I think I'd be more comfortable with Deprecated ("new implementations are no longer encouraged") as compared with Obsolete ("should no longer be [...] deployed").

Alright

So with 3 (+1) votes and 2 (-1) votes, we decide to change the state of XEP-0071: XHTML-IM to Deprecated

no other AOB? great

Sorry for the terminology confusion

thanks everybody

jcbrand, thanks for writing the minutes

Thanks all; sorry for the big contentious topic, I knew that would divisive when I first sent a mail to the list about it.

FWIW, Final XEPs have to have a replacement before they can be Deprecated. I don't believe that's true of Draft.

You're welcome

And any -1 prevents a XEP from advancing.

No?

Kev: no, 0001 says that advancement requires a simple majority

> A majority of Council members must vote +1 in order for a XEP to advance.

I think you're wrong.

Let me check.

Yes, the sentence before the one you quote "A XEP shall not be advanced to the next stage in the approval process so long as any Council Member continues to vote -1"

SamWhited, it also says it requires “rough consensus” on standards@, which it is far from imo.

It requires a majority to be +1, with no -1s.

Kev, I don't think Final has to be replaced to be Deprecated. It just says that if it is replaced, it'll become Deprecated.

dwd: I guess you could reasonably read it that way, yes.

Kev, true

I am now even more confused about the process. In that case, what are the reasons for -1s? Just that there is no other formatting XEP? What would make those concerns "addressed"? Discussion started? A SIG started? An experimental XEP submitted? Or does something have to get all the way to Final?

s/Final/Draft/

SamWhited, no.for me it would be enough to have an alternative experimental XEP

then we can say to people do that instead

Yes, same for me.

Something which could provide roughly as much interoperability as what we have currently.

So saying "don't do formatting in new implementations until something new is on the table" is not an acceptable recommendation for either of you?

(So not BMH, for which it’s explicitly a non-goal.)

i'll reply to the minutes that it's not actually deprecated

SamWhited, implementations do want formatting, and saying there is no way to do any kind of formatting atm is misleading at best, and harmful in general.

I disagree

I know. :)

But anyways, I'll start gathering use cases like I said in the beginning and we can keep this on the table. Although it sounds like it will be for the next council to finish again

Given that we're so close to the end of term

Thanks SamWhited.

jcbrand, will you cc standards too?

jcbrand, ta

Tobias: I always do, however I need to send from different mail accounts since I'm subscribed with different accounts. Who could I ask to change the email address with which I'm subscribed to council@xmpp.org?

Phew. Mailman admin?

don't know who that is though

I just reported my votes and Tobias’s on the trello.

ta