XSF Discussion - 2019-10-24

ralphm, https://github.com/xsf/xeps/pull/846 can this be added to the agenda. It should only be editor land it's a change in the schema but editors prefer to be on the safe side

(I couldn't add stuff on the trello board)

Yeah, I don't think this requires a board vote. Please make it so.

ralphm, I probably won't make the meeting later as I'll be travelling, but if I do it'll be mobile-only

ralphm, MattJ, I'm unfortunately in the same situation today :(

Me too

Nyco and me will hold the fortress!

Can Board follow up on the financial report? Apparently this was discussed at the end of last month (no minutes from that meeting and I was driving at the time) but never acted upon?

Daniel yes. It's still on our agenda.

from what I recall, we ended up debating if we'd like Peter to provide data from past years.

chat log here: https://logs.xmpp.org/xsf/2019-09-26?p=h#2019-09-26-ae8c71a4f222d202

From the looks of it there hasn't been a real conclusion and to my knowledge it hasn't been brought up again at other meetings. (despite 'still being on the agenda') - can I ask for it to be put on the real agenda?

If Board wants my input publishing current data (2018) is fine for now. And if it looks fishy people can ask for more?

It did not ever leave the agenda - it has been on my mind every meeting, but we simply didn't get around to doing it. One thing that I noted in the last meeting was that I think it'd be good to find a way to be more efficient in those meetings.

If he really has that stuff on paper publishing all of it seems a bit much. If it were just upload uploading some spreadsheet I'd be with you in asking for all of it

Daniel I'm curious why you bring this up. I have no problem with us releasing this information, but I personally have little interest in it (other than that I'd want to prevent us from being regarded as secretive / non-open). What's your motivation?

I brought it up *now* because today is board meeting and I had a minute of spare time to ask for it. It has been on my mind for a bit longer. I (as a scam member) would start to spend some money and I would like to see how much money we have, how much new money is coming in and what things we spend money on in the past. Just to get a grasp on what kind of money I'd be OK asking for

I've also heard rumors that we are not very good at actually collecting money from our sponsors

(sending invoices)

*would like to start spending

Daniel for SCAM, there's a yearly budget of 1000 USD that we can spend.

Right. I have however no grasp if this is a fair amount

Like what part of the annual budget is that

and yes, we've dropped the ball on doing invoicing / contacting sponsors in the past. That has even been discussed in this room, nothing about that is 'rumor' as far as I know. 🙂

Fair questions.

Like I said, the financial reporting never was on a backburner, even if we didn't get around to handle it in the last few meetings.

as for SCAM's budget - I'm not sure if we can, or should, qualify 'fair'. I'd have to check, but I think we've not spent any of that budget this year

(maybe some around the summit/fosdem, but that might also have come from last years budget)

so maybe it's more a question of: what do you want to do with the budget?

(And if we want to spend more than the budget, we can take that to board)

Guus, Daniel and I didn't know about it until not so long ago :p

"it" ?

The budget

SCAM's

Ok. Well, there is. 🙂

I've mentioned it just now in the SCAM MUC

(discussing this here so other non scam people can see this as well)

Yeah I've known about the scam budget for a couple of months /weeks now. That's fine

I could imagine myself asking scam to pay for sprint venues. We don't always have hacker spaces or something in the cities we want to do sprints in. And/or hackerspaces are not always ideal locations for reasons. Now if I were to ask for rent that can very quickly exceed the budget of 1k. However before I do I just want to know how much money for example 1k rent for 3 days is to the xsf

And then decide if I can justify asking for that

fwiw, the SCAM budget was requested as a means to not have to go to board for every expense, which (as you've found) can take a long time. It was not intended as any kind of hard limit that SCAM could spend.

Yeah I see the scam budget paying for day to day items such as merch and drinks at sprints

us wanting to spend numbers like that does warrant board involvement, I think.

But not for the entire sprint

right

That's fine asking board

I just don't know if it is a reasonable ask

but I understand your motivation to see the board financials.

And also even if Board decides people need to be able to understand the decision making

I'd think of it in terms of the Summit, which is the most important XSF-arranged event of the year, which we usually try to get sponsorship for all of and the budget comes in at less than that much usually (I think).

So I'd have thought there'd need to be a strong argument for why it was valuable to the XSF for spending that sort of amount.

It'd be nice to decouple budget from one important event - as such coupling makes it harder for other important events to be developed.

My point is that if we try to avoid spending much on the most important event we arrange, spending more on other events would need some significant justification. In my view.

Kev, there's an argument to be made that the sprints serve a good purpose - but I agree that for much money to be spend, we need to discuss that further.

Also, we can do many more events :)

We just don't have any idea if financials would even allow for this

They wouldn't, I think. We don't bring in that much from sponsorship, unless things have changed since I last remember.

I agree about all what Daniel said above fwiw

Kev: the issue is not that you think it would or not

I understand that people want to have the numbers to form their own opinion.

also, that might be a motivation to more actively attract funds.

Yes

To both

Kev also, I'm not sure if we are trying to 'avoid spending' money on the Summit, other than that we're careful to not overspend / overindulge, as we recognize that we depend greatly on donations of others?

To let money out, you want to make sure money will be comming in.

We seem to put a lot of emphasis on needing people to sponsor the room, the lunch, the dinner each year.

I think it might be hard to justify asking for all that sponsorship for the Summit, if the XSF is giving out significant amounts to run other, less well attended and core, events.

Kev I'm seeing that as us resorting to measures as we're doing a poor job of getting 'regular' / year round sponsors.

Yes, I agree with that.

Kev right, now I see your point

and I agree.

If we had buckets of cash, arranging more 'minor' events would seem like a good use of it. We just don't.

Opening the financials might allow others to come to the same conclusion.

(I almost typed 'collusion' there, I'm following the US news to closely)

Kev: even if that might be true, that's not an argument against revealing financials

No, which is why I didn't understand why Board were putting roadblocks in the way of doing so.

I don't think he ever made the argument to not reveal financials Pep?

I don't think board is putting in roadblocks, other than that they're slow to come to a conclusion

Guus: ah ok that's how I understood it

I thought "Oh, we can't just release them from now on, we have to keep them hidden until we can release a full history" was where it was.

Which I think is pointless.

we're basically debating: "should we release financials for the current year only, or also venture further down in the financial history?"

None in board argued to keep things hidden

I'd like to have the full history in the long run, but I'm happy with the current year as a start, better than nothing and that can already help

board being ineffective to come to a decision on wether to release current year only, or more history might have perceived as such. But, to be fair to board, I think they only discussed it once, in a meeting that was already running over time.

we seem to have at least 3 platinum sponsors? (given they have 'a page on the xsf') site. that's 30k

that doesn’t seem like we are super tight on cash

unless i don’t understand how expensive the summit is

which is why i would like to the the numbers at least for one year

more reasons to disclose those financials.

i don’t distrust people or anything. that's why for me 2018 would be fine

because i reckon all years will be similar

I'll try to push board to address this in todays meeting, but as you might have read, at least three members are, at best, attending from mobile devices.

also we have another 2-4k coming in from gsco?

Unsure - I don't wnat to speculate.

instead, I'll try to facilitate board coming to a conclusion on the publication of financials.

> Unsure - I don't wnat to speculate. Of course. Me neither. But that's what people will do if they can do quick math on one hand with the sponsors that are visible on the page and on the other hand have Kev saying that we don't even have enough money to pay for summit. But yes thank you Guus for following up with board on that

I think I have not disagreed with anyone here, other than that i don't believe that board is putting up hurdles for making financials public, other than that they're slow to come to a conclusion on this.

(Disclaimer for the log files: I currently serve on the Board)

Ok, I'm sorry if I was being unfair with the barriers comment.

It seemed to me there was a trivial decision to make (release for now on) that could be +1d in seconds, and then a more complicated second question that didn't need to block the first of asking Peter to reconstruct past records.

Kev, I think board didn't do an efficient job in discussing the topic, which bogged down everything. We should simply kick it in gear again, and finish the one discussion that we had, be done with it.

And I can understand how inactivity is easily misjudged as 'putting up hurdles'

https://www.indeed.com/viewjob?jk=113dde3cc7ce7284&tk=1dnufmkih40kn801&from=serp&vjs=3

Facebook Portal is built using XMPP ?!

ah no it's in the prefered qualifications

Maybe we could ask them officially to disclaim and, oh and also drop a few dollars :-°

pep. go for it

I mean "The XSF"

The XSF isn't a magical entity that out of itself produces questions 😃

too bad

do you want to draft something, maybe find a lead that we can contact?

I have unfortunately very little writing/communications skills (if it wasn't obvious enough will all I write here) :x

Unless you're happy with my sentence above

Very happy.

"Hey facebook. Give us monies"

Excellent. Now please find the right person at Facebook to send this to.

support@

We then have a concrete proposal, that the XSF can debate/improve upon.

postmaster@

looking for something, but I doubt this kind of information is publicly displayed anyway.

Having connections would certainly help a bit here

My first feedback would be: "it's my estimation that this proposal has little chance of success. Please improve it before sending it" 😉

That I could have guessed myself :)

Can we not get somebody to do that? As in, we have money, can we not spend money to get more money? That's usually how it works right.

We still need to find a point of contact. How many members do we have, they can be useful sometimes, maybe somebody knows something or someone that knows someone else

https://xmpp.org/about/xsf/members.html

That was rhetorical

> we've dropped the ball on doing invoicing / contacting sponsors in the past. Guus, I'd be interested in the extend to which this happened. Also, are there plans to still invoice the amount or has that already been done? To my (little) understanding of contracts, those companies would still legally be required to pay and the XSF already did their part (displaying the companies as sponsors)

I don't know exactly. Also, I don't think we should pressurize companies in paying things after we slipped up. That's likely to ensure no future sponsorships will happen.

Guus, if they are not paying, what are the sponsors good for?

from what I understand, it's us not invoicing some of them, not them not paying invoices

Also, this was quite a while ago, and was addressed.

So those that were not invoiced in the past are invoiced now? Didn't they ask why there were not invoiced before? Was that an actual board decision to not invoice them delayed or just continued laziness?

From what I recall, this involved sponsors where it was ambiguous if they'd silently extended their sponsorship with another year. Since this became known, we've purged the list of sponsors. Those that are still a sponsor, are invoiced. To my knowledge, there's no issue with payment.

(I've not actually checked the latter, but I have no indications that something is amiss in that respect)

So the actual problem wasn't sponsors not being invoiced but sponsors not being removed from the sponsor list after they were no longer a sponsor?

a bit of both, probably.

If we'd engaged them, they might have remained a sponsor for a few more years. SOme of the sponsors that we removed ceased to exist, so there's no point in following up with them.

(as in, company out of business)

makes sense 🙂

Guus, looking through archives: ProcessOne is a sponsor since 2015 (excluding May-July 2019), Tigase since end of 2016. Both of them seemingly expressed they want to be a sponsor in 2019. Was any of the two not invoiced for their sponsorship since they began being a sponsor (2015 or 2016 respectively)?

I don't know the details, not can I immediately find out.

😃

hello!

Board?

Board meeting time

I'm seeing three

👆✌

this morning, several people warned that they'd be on mobile, at best.

o/

I'm on a customer meeting, I'll follow with difficulties

Welcome at this edition of the XSF Board weekly meeting.

as usual, the agenda is driven by what's on our Trello board: https://trello.com/b/Dn6IQOu0/board-meetings

Previoulsy on "Board Meetings"... :)

1. Role call

:D

Greetings 🎩

I've seen everyone but ralphm (who this morning warned that he might not be available)

mentioning him in the hope he'll pop in...

While we wait: does anyone want to add something to the agenda?

Not me, thank you

Nor me

This morning, I've emailed about one of the items (financials) - I'd suggest we discuss that first.

I couldn't find time to read the thread, sorry

2. Minute taker

Would someone from the floor be so kind as to compose minutes?

ah, i forgot: last time, we discussed doing this live-ish, right?

as most of us are either on mobile, or busy, I"m thinking that's not feasible today?

right, I will propose a process, plus the tools for this

so for today, we could really use a minute taker, to not have to depend on chat logs

agree, mostly desktop solution

Today is not a good day for that due to mobiles :) but we can draft something out together after, checking the logs

sadly, no volunteers. If someone would like to do this retroactively, please.

In the interest of time, I'm continuing.

3. Publicize XSF Financials.

+1 ;)

Does someone want to comment on my earlier email?

> +1 ;) Nice

I'll read the thread before voting

As you and Seve already observed, we were all already in agreement about publication

off meeting

There was just the matter of whether we would ask Peter to go through his paper archives for historical stuff

Ok, I motion that we ask our Treasurer to make prepare for publication of a summary of XSF financials, to be published on the XMPP.org website, for the current fiscal year, as well as future years.

+1

+1

+1

Yep, I can do that.

I will - right, stpeter has been informed 🙂

:)

It will take me a few weeks, though.

while we have you - there was a debate last time, that revolved around us asking you to do this for past years

I don't think anyone expects this overnight, I'm fine with a few weeks (for the most recent data only)

The past years can be done, too. We once had this information on the website (several websites ago!), so the wayback machine could perhaps help us. But in any case I have all the information, I just need to compile it again.

There was an argument where people felt it unfair/unneeded to do this for past years, while others (well, me), suggested that it'd be helpful to have, and well, since you offered...

Is the amount of work involved significantly more, after you did the first year?

alternatively, we could circle back to this after you did the the most recent stuff only

If you decide to publish all can we prioritize 2018+current over 20 years of history

I've seen history on the website before, so I'm assuming that a large part of it is simply scraping old data?

(others, like myself, can help with that)

stpeter ?

We might have lost him. I think we can agree to have at least the most recent data. Let's circle back to having historical data too in a future meeting.

I'm not sure where on the site we used to have this information. It's possible I have it all on my computer, but I don't have time to look right now because I need to take my wife to the airport soon.

(By "all" I mean the earlier years - there will be a gap in the middle there somewhere.)

But prioritizing recent years makes sense.

understood. If we can at least a commitment that somewhere in the next few weeks we'll publish the most recent data, that'd be good.

Sure thing!

I'll take anything else as a bonus. I'd be happy to help you scrape stuff

Safe travels to your wife.

I'll visit the wayback machine over the weekend.

thanks!

4. Commitment for weeks ahead: Review of Roadmap page

Ralphs icon is with that, on Trello

We discussed this earlier, but I'm unsure what the current state or action is

crickets.

(We might also have the relevant files on an XSF machine, will check there too.)

let's park this for when Ralph is back

Ralph mentioned he would work on a list of possible topics for the Roadmap

5. Call for action: find candidates for the upcoming elections

Did we send out stuff through our comm channels?

I've not seen anything, but I've been away for a few days too.

I'm not aware, apart from the emails sent out by Alex

nyco do you happen to know?

Alas, we lost him to that meeting he warned us about

Does anyone have concrete plans to address this?

Have people in mind?

The current list of candidates is short: https://wiki.xmpp.org/web/Board_and_Council_Elections_2019

Especially for board

err, council, sorry.

I don't know of any person that could fit or be interested

We can tweet about it

Shall we draft an open invitation to be posted on our blog, and/or twitter?

sgtm

It would also make it easier for those of us in the community to link potential candidates to

(which implies it would be good for such a post to have an overview of responsibilities)

Yeah. We could take Alex's default text as a template - I'd love to expand on it a little though

Yeah, that sounds good MattJ

Does anyone have a concrete text in mind, or shall I draft a suggestion?

I don't, but happy to review a draft

ok

As it's time

6 AOB

Anyone?

None

None

If I'm not mistaking, DST switchover is upon us, this weekend. Board members are all in the same switchover region, but others might be affected

(just in case you're in the US and REALLY want to be here for the next meeting?)

7. Date/Time of next

+1w

All good!

<the end>

Thanks

Thanks Guus

Thank you Guus

MattJ _rough_ draft sent by mail ✏

> Guus> This morning, I've emailed about one of the items (financials) - I'd suggest we discuss that first. on board@ that's not public?

Correct

I don't believe it contained anything that wasn't discussed in the room today or in previous meetings

financial summary 2001-mid 2008: https://web.archive.org/web/20150315021548/http://xmpp.org/about-xmpp/xsf/xsf-financial-summary/

Thanks larma / ping stpeter

larma: thanks for that!

stpeter, "(By "all" I mean the earlier years - there will be a gap in the middle there somewhere.)"

What did this mean

(I'm compiling minutes)

Thanks pep.

I read that as that he might have all data that he compiled earlier still on his computer somehwere, but that he didn't compile data from all available rough data

We have records for the early years. I will create financial summaries for the recent years, starting from 2019, but I will need to fill in 2008-2018 because https://web.archive.org/web/20150315021548/http://xmpp.org/about-xmpp/xsf/xsf-financial-summary/ indicates that I last did this in 2008.

(Well, we have *records* for all the years, but they're in a fireproof safe at my house.)

stpeter, I see

So it'll just take time but data is present

somewhere

Yes, the data just needs to be compiled - i.e., I need to look at all the expense reports and add things up. Not that hard.

k

OK, now I actually need to go, I will check in later.

Minutes sent.

https://github.com/xsf/xmpp.org/pull/626

"only a few hours a week" hmm maybe this should be removed

I'm out for the day

Or at least a couple of hours

Please leave comments in the PR

k

What types of invites do clients currently use with MUC? Do any clients do direct invitations?

larma, FYI, we already depend on the unicode standard for our wire format, see also the most recent discussion at the IETF's XMPP WG mailing list

also, I think that if you go for bytes or codepoints than sender (and ideally also recipient) should normalize the unicode string frist

yes gajim in some circumstances

MattJ: I found a handful of mediated invites in my archive, no direct.

about the sticker discussion

byte seems very weird, i have to decode the stanza from the wire

then get the body data

then reencode it? split it up into parts

I (posted to the list that I) would go for codepoints.

decode the single parts again

Everything except JavaScript should be able to deal with that.

Wait, what

Those were using XEP-0249 syntax but were sent from MUCs

also whole discussion sounds like someone *wants* to find problems

Human language is complicated.

Human is complicated

lets just use codepoints and be finished with it, and questions like what if a graphem consists of mutliple codepoints and reference links into it

yeah what then? nothing then, i apply bold to that and if what comes out is garbage

i open a issue with the client that sent me that garbage

but thats about it

yeah

Bugs will probably happen regardless

I am undecided between codepoints and extended grapheme clusters, but I believe that if we use codepoints we should normalize the strings

Alternatively, pick a magic marker code point and replace that.

cause otherwhise the codepoints may be different. Although this may be unlikely, I wouldn't risk it and I believe most platforms provide unicode normalization APIs, so it should be easy

or bring back xhtml-im and use <img>

the rumors that xhtml-im is gone and needs to be brought back are exaggerated ;)

`git revert 48593ca5b8659553c02ce7b9e12fa7660436d9bc 4c5db548251a91d8e6c55b1add84f92765d172a3` and pretend nothing happened

flow, I don't think the IETF mailing list issue is a real issue. Basically it says: PRECIS doesn't define the Unicode version to be used, so we are fucked. It's true that this is an issue in the standard, but also it shouldn't really matter in practice as long as you follow the rule of being strict on what you control and liberal what others send you. The issue of counting characters is much more imminent and will have real impact if not handled properly *now*. People are already using markup/reference XEP and do basically random stuff when there is bodys multi codepoint graphemes in place. Regarding normalizing: Clients probably should normalize before sending (and before determining markup/reference positions), however normalizing is Unicode version specific behavior, so a second normalizing by a different entity (using a different unicode version) might end up and further changes and thus this entity would be required to update the markup/reference positions. So it's probably better to just not do it when receiving or forwarding if you don't want to deal with that. Also note that normalizing might actually break things if the sender uses a newer unicode version for normalization.

> as long as you follow the rule of being strict on what you control and liberal what others send you.

you mean the thing you *shouldn't* do ? https://tools.ietf.org/html/draft-iab-protocol-maintenance-03

aka draft-postel-was-wrong or somesuch

I liked the original working title better hehe https://tools.ietf.org/html/draft-thomson-postel-was-wrong

+1 to Zash for having the best memory

larma, I don't think we are fucked (mostly because I think what PRECIS does is sensible). All we need is agility regarding the supported Unicode standard in runtimes or libraries. In fact, chances are high that your system has a sufficiently recent unicode database in /usr/share/unicode(-data)

And what the draft-iab-protocol-maintenance I-D says

Wait didn't that get RFC'd?

Huh, not yet I guess

flow, sure for sending this is a sane thing to do, but for receiving you should be more liberal. After all your cannot ensure OS and/or client are of recent version.

be liberal or fail loudly

but also don't fail.

> chances are high that your system has a sufficiently recent unicode database in /usr/share/unicode(-data)

Any Redhat ever might surprise you!

Mmmmmmmm Enterprise

moparisthebest, at least fedora has a unicode-ucd

larma, I am not sure if you can be liberal, or, at least, if it wouldn't cause more problems in the long run. Being liberal in what to accept often sounds like a nice easy solution, but it sometimes causes unforeseen issues and in general causes the issues described in the I-D

Maybe the real answer is "it depends"

as so often

I don't know about unicode, but up to date redhat 7 has a curl that was released Feb 6 2013, and up to date still supported redhat 6 has a curl that was released in Nov 4 2009

unicode is surely just as bad if not worse

What implementations could do is to use an old unicode version when verifying and "generating" JIDs, to ensure maximum interoperability

strict mode when creating users and such?

wouldn't that disallow things allowed by newer unicode versions?

yes, that's the point

moparisthebest, yes, but if you prevent the user from even generating such JIDs you achieve maximum interop

to ensure maximum interoperability maybe impls should just limit everyone to ASCII

wanna create an new MUC? make sure to use unicode 6.0 (or so) to prepare and enforce the JID that you send to the MUC service

and now you're on the unicode upgrade threadmill! enjoy

them users gonna want those new codepoints, better upgrade

it's software and protocols, we are always in the upgrade threadmill

Is CentOS 6 dead yet?

I don't think so, redhat 6 isn't

https://cerdale.zash.se/upload/ilURA86kFBb66daT/image.png

There was something about ligatures but I don't remember the exact details.

Zash, ligatures are a font only thing that affects displaying but not anything the toolkit is doing, so you can still select half of a ligature. Depending on toolkit, the selecting thing can already change how ligatures are displayed. I don't remember which, but I know I have once seen a messenger that manually rendered the cursor in the input field and this broke ligatures if the cursor was between the two characters that make up a ligature. And this actually can affect the two chars length (if they are shorter as a ligature) which could cause the multiline text field to wrap differently. Long story short: moving the cursor through the text field would actually cause the text in it to jump around.