XMPP Council - 2019-01-16

Do we have agendums for today?

if i have a question regarding the state of a xep, where should i ask it?

either the xsf MUC or the standards email list are good venues

status

oli, xmpp:xsf@muc.xmpp.org?join is a good first approach

good

Uh, I’m going to a conference tonight, it starts at 6pm at about 1h from where I live, which is exactly when our meeting starts. :x

I’ll go there earlier, and will look for wifi.

If you don’t see me by 5pm, assume I’m dead^Wnot somewhere wifi exists.

This time there is no Agenda.

I'm principally available and probably also full-time-available at Meeting start.

Due to an unforeseen infinite loop in JabberFR’s Prosody, my plan to leave early got foiled; I won’t be present for the meeting, sorry. :(

Link Mauve: what a convenient excuse! 😁 Have a nice evening!

Sorry folks, 1600 fails for me. Just got a short-term appointment.

here I am

And me.

But no Dave, Link Mauve or Ge0rG, so no quorum, I think.

do we need quorum to start possibly on-list voting?

Technically, yes.

darn

Give Dave until 5 past, otherwise call it.

mildly frustrated because it took quite some effort to be here on time :/

Ooops, sorry, didn't notice the time.

1) Roll Call

I'm here, obviously.

That's us, isn't it?

Ge0rG and Link Mauve excused themselves on short notice, yes

1) Agenda Bashing

No.

2) Agenda Bashing

we have a protoxep inbox

We have a protoxep?

did I fail to announce it?

that’d explain things

no, I did announce it

https://xmpp.org/extensions/inbox/hash-recommendations.html

the open LCs aren’t due yet

and afaik there are no other points from the editor

so only the protoxep

Indeed. I entirely missed that one, somehow.

But anyway:

3) Items for a Vote

(I am aware of at least one copy/paste issue in the protoxep, which I can fix before we vote on it if that’s desired)

(some reference to a Version 1 of the hash namespace which doesn’t make sense after the split)

a) Cryptographic Hash Function Recommendations for XMPP Abstract: This document provides recommendations for the use of cryptographic hash functions in XMPP protocol extensions. URL: https://xmpp.org/extensions/inbox/hash-recommendations.html

+1 obviously

I'm on-list, obviously, having not seen this one. But I'll place a +0 on it for now.

Kev, any vote from you on this one?

Sorry, on-list.

Fair enough.

4) Outstanding Votes

Looks like the only two outstanding votes are from Ge0rG and Link Mauve.

5) Next Meeting

AOB

Same time next week?

jonas’, Noted.

wfm

WFM, AFAIK.

6) AOB

jonas’, You're up.

a) Secretary bot open for betatesting in xmpp:council@conference.zombofant.net?join

Why not in here?

I’d like your feedback, what you wish it can do, what it does wrong etc.

Kev, avoid spam in this very good signal/noise room

but I can put it here, too, if that makes sense?

jonas’, Is the source up anywhere?

not yet

can change that in a minute though

(I also have another AOB)

jonas’, Personally, I'm happy to spin it up here. MVP is very M, so we might was well try it.

ok

will put it on here after I moved it to some machine which isn’t shut down every day for hours :)

jonas’: Much more likely it'll get tested live, I imagine.

If it's broken and starts spamming and you're not around to fix it, banning it is trivial.

true

https://github.com/horazont/councilbot

source ^

Awesome.

jonas’, Next AOB?

b) Deferred XEPs

the other day in xsf@, I came up with the idea that we should look at a random(*) deferred XEP each week

to kind of sort where they stand, because it’s often not obvious

I see three states: needs more work, should be LC’d with prospect of advancement, should LC’d with prospect of being dropped

I think we have quite a few in the last two states, which is something council can do something about

so... I’d draw a random XEP each week, announce it here or somewhere, and everyone takes a short look and we figure out what to do with it?

I think that presupposes that we should avoid having XEPs quietly languish in Deferred.

I think there’s some merit of cleaning up our XEP list, because it has been reported as being overwhelming

I don't think it presupposes that.

But what's the difference between a XEP list that has lots of deferred and a XEP list that has lots of Rejected?

I think it just presupposes that there might be XEPs in deferred that shouldn't be.

dwd, I’m hoping we’ll have more in Draft actually

Deferred <-> Rejected I’m not so worried about, Deferred <-> Draft is more relevant

however, I understand that we all are busy people, and you’d prefer to distribute the workload on the authors

Oh, no, there was a line in there that presupposed they shouldn't be deferred. Apologies. I disagree with that notion.

(that was Ge0rGs stance, IIRC)

I'm fine with looking at one a week, but the possible (and likely) outcome is "yep, should be in deferred".

OK. I'm all in favour of checking through XEPs to see what can be usefully advanced. I'm not sure that a random scan is ideal.

dwd, I’m open for any other mode

I don't see a problem with just starting with oldest deferred and moving forwards.

oldest by xep number or oldest by last update?

I feel that the oldest Deferred XEP is least likely to be advanced. We might be better off looking at most recent Deferred (or oldest Draft).

dwd: Least likely to be advanced, most likely to be deprecated/rejected, I think.

Anyway, I am not high-F on the mechanism, other than that I think sequential (by number) is sensible.

Direction matters to me less.

Kev, Potentially. I'm a lot more interested in advancing stuff that needs to be advanmced, though.

I do like a good advanmcement.

Kev, Yes - I'm low-F on deprecating deferred-and-useless XEPs, but higher-F on advancing deferred-but-worthy XEPs

But anyway, I'm happy to give it a go.

Any other AOB?

not from me

Newp.

OK. In that case:

7) Ite, Meeting Est

thanks dwd

Thanks all.

Diolch, pawb.

And once again, sorry for being late.

before you all vanish, can somebody give councilbot@dreckshal.de moderator permissions so that it can see the occupants JIDs? that’s required for secure voting.

Hi. Have I missed something? 😁

Ge0rG, yes, the meeting ;-)

ping

jonas’, pong

Secretary, create vote on something

I guessed that much, no privileges yet :)

jonas’: darn!

I think the Secretary syntax urgently needs shortcuts

Ge0rG, have you read the syntax manual? ✏

jonas’: I have. I want *shorter* commands, not longer

right, I wasn’t sure whether you were going off my verbose backlog in the other room

make concrete proposals, I’ll implement them

!help

Kev, it only listens to stuff prepended with it’s name

also, here: https://github.com/horazont/councilbot/blob/master/docs/patterns.rst

Secretary: help

jonas’: !help !+1 !-1 !list

also fuzzy matching of <the subject>

Ge0rG, it does match fuzzy on the subject ;-)

(and if it matched unexpectedly, you can LMC it)

also, I need to cycle it so that it picks up on the newly visible JIDs

Secretary, help

jonas’, TBD

Secretary: help

;-) ✏

Ge0rG, TBD

ah, it accepts `:` as well

it could link to that rst now

jonas’: is that rst part of the github repo?

link to the formatted version then, please

yes

and yes

Secretary, help

jonas’, TBD

ey

Secretary, help

jonas’, https://github.com/horazont/councilbot/blob/master/docs/patterns.rst

so can we please have bang commands

on it

Secretary, create poll new: the opposite of old

Ge0rG, created poll on new: the opposite of old. Expires on 2019-01-30

Secretary: vote -1 on new: the opposite of old: it's not actually new

Ge0rG, sorry, I do not know which vote you mean.

Secretary: vote -1 on new: the opposite of old

yeah

Ge0rG, sorry, I do not know which vote you mean.

that.

Secretary: vote -1 on new: not new?

Ge0rG, sorry, I do not know which vote you mean.

it only matches on the part in front of the colon

the other part is irrelevant for the match

Secretary, vote -1 on new the opposite of old: because I can do it nevertheless ✏

jonas’, I recorded your vote of -1 on new: the opposite of old: because I can do it nevertheless

Ge0rG, ^

wait, what?

fuzzy match is fuzzy :)

so I guess this isn’t too much of an issue after all?

Is there a reason not to cycle through in normal meeting style?

The way the old meetingbot did, but with the addition of remembering votes?

Kev, I intend to make it so that a newly created poll is current for 30 minutes or until a new poll is created

and then vote commands without subject words would match that poll

(the idea is that you can vote using the bot outside the meeting, too)

So you create them in the week before Council, and then !next a few times to get through the meeting, with it detecting votes each time?

more like chair creates the polls in the meeting when moving to the next agendum

(since expiration is tied to creation time)

It's not entirely clear to me how that's helping, then?

how what is helping in which regard?

Keep track of the agenda?

my goal was to keep track of votes, not of the (pre-meeting) agenda

Oh, ok.

and opt-in reminders to pending votes and stuff

Any chance we could have it do agenda management? It used to be great being able to just !agenda (or whatever, I forget) to see what the agenda was looking like being for the next meeting.

sure

no reason that couldn’t be implemented, too. I’d keep it orthogonal to keep the data model simple though.

(and the interaction model)

Secretary, please do I want to please

Back when we used to have this before it was so nice as chair to just !next through the meeting and have all the agenda passed out for you.

somebody read the source.

well done, dwd, you broke it,.

Secretary, I do want to please nevermind

dwd, as if it never happened

not fatally, but you made it complain

("why nevermind?" -> "for use with LMC")

Secretary, nevermind ✏

jonas’, as if it never happened ✏

Secretary, list open polls

jonas’, there is 1 open poll new: the opposite of old (due in 2 weeks, on 2019-01-30)

Secretary, Please show me all the open polls

dwd, there is 1 open poll new: the opposite of old (due in 2 weeks, on 2019-01-30)

I actually rather like the natural-language interface.

(I somehow expected that dwd would be the one playing with the long syntax, I do not know why)

I plan to keep it in addition to the !-commands

Odd. I really can't find that email about the hash stuff.

Oh, there we go.

Kev, the protoxep or the discussion about splitting?

Kev, ah, it doesn’t like your JID

I’ll fix that

Kev, should be good now

nice

Secretary: create poll Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html ✏

it doesn’t like when the first message it sees from a member is an LMC

stupid assumptions!!k

Secretary: create poll Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html

Kev, created poll on Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html. Expires on 2019-01-30

jonas’, Something that would be nice is if it could take "create a new poll on https://xmpp.org/extensions/inbox/hash-recommendations.html" and enrich it with the title and abstract

dwd, on the list, yes

also for github PRs

+1

Indeed.

Secretary: +1 ✏

Kev, sorry, I did not understand that.

Secretary, I vote +1 on Accept to Experimental

jonas’, sorry, I do not know which vote you mean.

Secretary, I vote +1 on Accept Cryptographic Hash Function Recommendations to Experimental

Secretary: +1

jonas’, I recorded your vote of +1 on Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html: (no comment)

Kev, sorry, I did not understand that.

Kev, the context stuff isn’t implemented yet

Secretary vote +1

Ah, ok.

Secretary: create poll Accept "Some dummy XEP" to Experimental: https://xmpp.org/extensions/inbox/dummy-recommendations.html

Kev, created poll on Accept "Some dummy XEP" to Experimental: https://xmpp.org/extensions/inbox/dummy-recommendations.html. Expires on 2019-01-30

A vote number might be a helpful shortcut. Or even if we could create shortcuts, so "vote +1 on hash-recommendations" would work.

Secretary vote +1 on accept

A substring search might actually be good enough.

dwd, the match is already fuzzy, but it needs some confidence to be accepted

this is particularly bad with long titles and URLs :/

Secretary, vote +1 on dummy

dwd, sorry, I do not know which vote you mean.

Secretary, vote +1 on some dummy xep

dwd, sorry, I do not know which vote you mean.

(something which matches if it’s the only thing where the word occurs would be good as a fallback, indeed)

Secretary: vote +10000 on whatever

Secretary, you are useless.

dwd, sorry, I did not understand that.

jonas’: I think your concern about SNR with the bot in here was completely unfounded.

Secretary, You have supplied the proof of my assertion

dwd, sorry, I did not understand that.

Zash, DEBUG:main.client.bot.CouncilBot:ignoring message from non-member zash@[…]

This backlog is priceless.

Kev, COMPLETELY UNFOUNDED it was

jonas’: what about using `*tag*` within the title as an optional implicit shortcut?

not sure which part of your message is intended to be markup and which is intended to be verbatim

#in-body-markup-rules

Shall I reencode in XEP syntax?

maybe just without markup

if there’s any

EBNF?

jonas’: if the title contains a word surrounded by asterisks, use that as the shortcut

ah

I'd suggest square brackets for the markup reason, actually.

dwd, +1

also because square brackets

Secretary: vote +1 on Square brackets

Ge0rG, sorry, I do not know which vote you mean.

heh

dwd, sorry, I did not understand that.

Secretary, Please can you create a vote on square brackets? ✏

dwd, sorry, I did not understand that.

:-(

(it doesn’t ignore "can")

jonas’, It needs to ignore me more.

(or "you" for that matter)

yes

Secretary: please create a misunderstanding

Ge0rG, sorry, I did not understand that.

jonas’, sorry, I did not understand that.

Secretary, can you please create a vote on the square brackets? ✏

jonas’, created poll on square brackets. Expires on 2019-01-30

dwd, there you go :)

so I understand that you folks want: - agenda - square bracket tags for easier voting - fetching metadata from known URLs

Square brackets when creating the vote, not when actually voting, please. Also the obvious interaction between agenda and votes

Ge0rG, sure

Ge0rG, https://github.com/horazont/councilbot/issues/1

and https://github.com/horazont/councilbot/issues/3

jonas’: awesome!

so I hvae something to do for tonight :)

Kev, (or anyone) do you have a reference or example usage of that old agenda bot? it isn’t clear to me, for example, how the order of the agenda would be determined -- by time of addition seems a bit meh, and adding a text interface to sort it seems meh too ✏

(ad-hoc with text-multi or something comes to mind, but I haven’t implemented the server-side of XEP-0050 in aioxmpp yet)

Secretary, please list the votes on accept cryptographic hash function recommendations as experimental

jonas’, poll on Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html is open. The poll is failing. Kev has not voted Ge0rG has not voted dwd has not voted jonas’ has voted +1 without further comment Link Mauve has not voted

that this pings everyone is probably not good, does anyone have a suggestion how to fix it? ^

I have seen a bot adding an underscore right between the first and the second chars of the nick

it "works" most of the time

mathieui, that sounds like a good think

when does it not work?

well, it works in the normal case

which is the non-normal case?

when people take nicks including other people’s nick, it breaks down

but that won’t be the case here ✏

ah, so when someone would call themselves K_ev, that’s right

do we have something which strongly allowed in resourceprep, doesn’t look terrible and would confuse typical highlight detections?

Secretary, please show the votes on accept cryptographic hash function recommendations as experimental

jonas’, poll on Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html is open. The poll is failing. j_onas’ has voted +1 without further comment dwd has not voted Ge0rG has not voted Link Mauve has not voted Kev has not voted

that didn’t work

almost

Secretary, please show the votes on accept cryptographic hash function recommendations as experimental

jonas’, poll on Accept "Cryptographic Hash Function Recommendations for XMPP" to Experimental: https://xmpp.org/extensions/inbox/hash-recommendations.html is open. The poll is failing. j_onas’ has voted +1 without further comment L_ink Mauve has not voted d_wd has not voted G_e0rG has not voted K_ev has not voted

better

and, another "yay for in-band signalling"

Why don't I see anything from secretary at all in yaxim?

https://upload.yax.im/upload/-gdFfyemG-YPSYCF/20190116_220730.jpg

Must be the @xml:lang on the body. Meh.

Now I wonder if this is a new smack4 "feature" or if it has been broken forever

jonas’: you could use this point · instead of _

j·onas’: