XMPP Council - 2010-10-18

Evening Ralph.

Thanks stpeter :)

waqas: :)

Hopefully my connection will survive for an hour.

The radar page in the topic is months out of date. Is there an agenda somewhere?

On the Council list.

Hi all, a possible agenda for tonight follows - please read and bash. 1) Roll call 2) Agenda bashing 3) XEP-220 (Dialback) This is Experimental - shall we issue a last call so we can move it up to Draft? 4) XEP-201 (Threads) Last call finished a while back on this. Are we ready to vote up to Draft? 5) Date of next meeting 6) Any other business

I'm on a train. Let's hope I continue to have a connection.

just don't try to vote when you're in a tunnel. :)

let's make it quick then. There shouldn't be a tunnel in the next half hour

Two minutes yet.

I don't see Matt online.

Haven't seen him today.

yum, lunch

dito, haven't seen him either

Righty, so.

1) Roll call.

here

Kev, Ralph, Nathan, Remko here.

2) Agenda bashing.

Anyone?

no

none

none..looks like the radar is pretty current

Fritzy: amazingly

3) XEP-220 (Dialback) This is Experimental - shall we issue a last call so we can move it up to Draft?

+1

+1

It seems to nicely coincide with me having to fix some issues in Wokkel around that.

Remko?

:)

ralphm: :)

kev: i haven't read it :\

Kev: i'll reply on list

It's dialback, it used to be in 3920 :)

But ok.

4) XEP-201 (Threads) Last call finished a while back on this. Are we ready to vote up to Draft?

remko: you can read it during last call!

oh ok

if it's the same thing, +1 :)

The last thing that happened in the last call feedback on 201 was Peter saying he was going to go research IMAP.

I haven't seen much feedback on 201.

kev: +1 on the last call for dialback, i misread it as an 'advance to draft'

"It seems that would be RFC 5322: https://datatracker.ietf.org/doc/rfc5256/ Thanks for the pointer, I'll give that a read."

stpeter: did you get anywhere with that?

XEP-0201 is informational

Kev: I added a reference to some IMAP specs, but they are not directly applicable

ralphm: Right, not Draft.

as far as I can see, anyway

stpeter: So you're happy with it as-is?

If so, does anyone object to us scheduling a vote for Active for the next meeting?

no

No objection. I think it's a good informational spec.

has 201 been used much?

Kev: I think it's fine as-is -- we might need to revise it again in the future after more people implement threads

and do recall that it was connected originally with some Ian-isms like 155 and 116 :)

but I think we've scrubbed most of that

I'm planning on making it the default behavior in a future release of Sleek.

Fritzy: ah, nice

Ok, let's vote on that next time.

+7

AOB? I have a query about voting of Council/Board that I was going to bring up on list today, but was busy.

Or, next meeting rather.

Same time next week?

wfm

same time is ok

hooray

So, elections.

The current bylaws are that the Council has an upper bound in size set by the members (currently 5), and that Council is made up of those people with the most votes, and an affirmative vote by the majority of voters, up to that limit.

right

oh no, voting rules again?!?

stpeter: I think the rules are right, but memberbot is incorrect.

I think we changed it so that members shouldn't be limited to voting for five people.

ahhhhh

Revote!

Because with someone needing a affirmative majority, we could well end up without a full Council.

ralphm: :P

e.g. 8 applicants, 2 people get a vote from all members, the other 6 share the votes just less than equally, and we only have two people on Council.

seems unlikely, although possible

It's not that infeasible if we had e.g. 3 strong candidates and 5 less strong - not suggesting that we do.

I was wondering what other people thought.

I'm all in favor of someting re-writing memberbot :)

Are the memberbot sources public by the way?

I don't think this particular change needs a rewrite, although I'm all in favour of someone doing so.

waqas: they should be -- but in fact it would be better for someone to start fresh because the code is not very good

I have been writing some bots lately (for statistics), and was going to do one for filling forms.

Tobias has written such a thing for gsoc a couple of years ago, and Gislan (I think) wrote a memberbot plugin for sleekbot to replace the current memberbot.

Kev: what do you suggest?

I could find some time to do a rewrite this week, I think.

I don't think we can rewrite the bot in mid-stream

Kev: and i even heard back from a single user from australia :P

Mine would be a plugin for Riddim (bot based on Prosody sources)

stpeter: no, I think we'd need to cancel the current votes and redo them.

erk

waqas: I'm opposed to using Riddim unless someone commits to making it work :)

I run Riddim in a couple of places, and it's great if you're willing to kill it and restart it when it falls over, but that's pretty often.

Kev: but members still have only 5 votes -- the difference is in how they're counted

stpeter: Why do they only have five votes, though?

so the problem is that you need to decide for whom you want to vote before you start

Kev: We were planning on making packages, and it'll probably get plugin reload without restart soon, same as Prosody.

This isn't a bylaws thing, justsomething we've traditionally done.

or, if you don't like how things went, ask Alexander to throw out your votes and start again

um

The problem isn't memberbot, it's that we're artificially limiting members to five votes.

well, we're electing 5 people -- why would you get 8 votes?

we're asking you to affirmatively state that you want these 5 people on the Council

which doesn't really happen

Right, but each person must have >50% people voting affirmatively on them.

if we're going to change the voting system, we might move to preference voting or somesuch

why not ask which people you would want on council, be it more or less than 5

Unless the names are randomized for each voter, the first five probably get significantly more votes :/

i thought they were randomized these days

They are.

you look at the list of candidates, decide "I was Alice, Bob, Charlie, David, and Elizabeth" and vote yes for them, no for everyone else

waqas: they are randomized

stpeter: Yes, I understand how to vote :)

I always decide before I start voting what my votes will be

Kev: so what's the problem?

I don't really see an issue

I don't see anything to change here

at least not for this year

Let's say we have 12 people standing next year.

and the votes go something like (making it up, they won't add up right)

if folks want to have a long debate about voting processes for 2011, I'm all for it

45% 45% 45% 45% 45% 30% 30% 30% 30% 30% 20% 20%

I suggest fixing the voting system when we get into trouble

We end up with no council

it's not like we can't fix stuff after it breaks down

I need to get going.

Or no board, I think, because Board elections have the same 50% requirement in the bylaws iirc.

Anyway, as long as people are convinced there isn't a problem, I'll go sit quietly in the corner.

:-D

Kev: I suggest we burn that bridge when we come to it

Ok.

AOAOB?

I see the concern (now)

I think the bylaws have a provision for solving unforseen issues by ubermajority

nod

Ok, no other business then.

Kev: I suggest you hold your breath, then bring this up after the current round of elections :)

In that case, Thanks all, see you next week.

thanks

I'll do some minutes tomorrow morning probably.

somehow I had not seen the possibility of this :(

anyway, for Council we'll issue a Last Call for comments on XEP-0220 and expect some more votes on XEP-0201

stpeter: Well, we'll LC on 220, and Council will vote on 201 next week.

ah

WFM

won't next week's meeting be the last one of this term?

Probably, I've not looked at dates.

ah, so it the 25th

yes

an hour after council meeting starts

I just asked for clarification about the time

19:00 UTC?

For Council or Members?

Council is 18:00 UTC next week.

Members

it says so in the announcment

I missed the announcement :P

I never received an announcement.

I only knew voting had started because memberbot appeared online.

there were some emails on the list

I had the initial meeting announcement back on Sept 20th

But not one to say proxy voting had started.

The initial announcement said 1900

it was sent on sept 9

eh, 8

ralphm: What was?

Ah, that mail.

the meeting notice

Yes.

Not the voting announcement :)

ok

calendar updated

Thanks.

it also noticed voting

starting at oct. 1

Yep. Which didn't happen, as far as I know.

Kev: I'd prefer to debate voting procedures after this round -- the scenario you raise becomes more likely as we have more candidates, but there is also the possibility that we have 5 candidates for 5 positions and one of them is deeply unpopular (receives only 20% of the vote or whatever) and then the person would be voted in despite receiving all those no votes

I'm not proposing changing the bylaws, which I think are right.

I think it's right that if 90% of membership think someone is unsuitable that they won't get in.

I thought perhaps you were suggesting that we remove "with the proviso that no individual receiving less than a majority of votes cast shall be elected"

The problem I see is that membership aren't given the chance to say whether each candidate is suitable or not :)

No, I believe that is absolutely right to have.

Membership are able to express up to five people who they think are suitable, but in the current voting system (not the bylaws, just how we count votes) it says nothing about their opinion of the other X people standing.

ah

sure

so we need preference voting :)

rank them all from 1 to 8 or whatever

I mean, all of these folks would be qualified

or 'abstain' per candidate, next to yes and no

I don't think that's even true.

it's hard to vote "no" on any of them

I think ideally we have a two vote system.

Or two-track, rather.

http://en.wikipedia.org/wiki/Condorcet_method

Time's up. Eindhoven

One track for voting for each member if they're suitable or not (this should be >50% for any member of Council), and one track for expressing preference.

ralphm: tot ziens!

too complicated

And no, I've frequently believed people have stood for Board/Council who weren't qualified :)

well sure :)

So simply the X most popular isn't a reasonable metric, because that could pick people who everyone unanimously believes are unqualified, given a scarcity of competent candidates.

This is terribly unlikely to be the case this year.

so we either live with the consequences and hope for better candidates next time, or modify the voting process

But then next year, when only 5 people stand, one of whom joined the XSF because it was cool and hasn't ever read or understood a XEP...

heh

brb