XSF Discussion - 2017-09-20

mathieui thanks!

Did I get this right: Psi(+) does not support OMEMO because this kind of encryption is forbitten in Russia?

https://github.com/psi-im/plugins/issues/10

*forbidden

no

What?

rion is full of shit there

So the real reason is more the lack of willing developers ...

yes

ack

hi

Board meeting?

ETA -5 min?

I got Ralphs "hi" after my last message, but according to Conversations, it was sent more minutes ago?

Guus, according to gajim his "hi' was 5 seconds before your "Board meeting?" message

My client said 5 minutes

Weird

I sent it at 18:05:06

This week on the XSF Board; time travel.

(CEST)

Guus: your message was delayed, so probably you were not properly synced to the MUC at that time.

my conversations and gajim both agree with ralphm's time, I think your client is messed up Guus :)

I'm here, not sure who else is…? Also, yeah my client's doing something odd

Still, board meeting?

I haven't detected quorum yet

I'm here

17 minutes in

Are there any pressing matters this week?

I don't know how pressing it is - Guus requested approval of his logo fix

My issues don't have a deadline, so feel free to skip board meetings until the next re-election.

Ge0rG: thanks for that

ralphm: I'm sorry. Venting at the present board members about the absence of the non-present ones is highly inappropriate.

Ge0rG: Well, I've been absent quite a few times, too. Doing volunteer work is hard some times, and so is attending (weekly) meetings. That said, I don't feel that we are doing a bad job in general. Most of the work in this organisation is around protocols and that's why we have Editors and Council.

FWIW, I feel like we're doing a bad job

ralphm: I think I'm repeating another person's point, but it's up to the board to decide how often to meet, and at which time.

I've written at length about this, without response. Cannot elaborate again now, driving.

MattJ: ok

There have been several times where people have needed things from the board and the board did not show up for several weeks in a row. That feels like a problem to me.

(or rather, where a small subset of the board didn't show up; the people in this room tend to be the ones that are on top of things I think)

The last board meeting was very intense and very productive, though. It didn't cover all the points, but still.

Agreed

SamWhited: sure, attendance can be a bit spotty, and from experience especially around summer leading into the autumn.

Last week was indeed nicely productive

Theta now is quorum, please meet?

0. Welcome + Agenda

Hi all

Hi ralphm

Ahoy

I understand one of the topics is Guus' request to change the XMPP logo.

anything else pressing?

See trello

Guus: thank you, but that's not what I asked

There is a trademark question on Trello we didn't get to last week

Ok

and the SPAM WG question

Let's see how far we get.

1. XMPP Logo

+1 to Guus's changes

I know there's been some differing opinions on this and if Board can decide on this.

Kev is -1, most others that expressef opinion +1

I'm OK with Guus' changes

Wasn't the logo initially voted upon by the membership?

I'm suggesting a small improvement, not a new logo.

Guus: I don't see the difference, to be honest, procedurally, I mean

I don't see a need to change the logo myself, but I'd like to make sure that we do it right.

ralphm: original vote if any was between different styles

the email went out to the membership, presumably everyone who cared already commented?

moparisthebest: that's not the way things work in foundations in general. Either Board can decide, or the membership in a general meeting.

In the mean time, I've been trying to find background in how this logo came to be, but haven't yet found the details.

ralphm: it is in my pr

Including feedback of original designer

if you don't know the right way to do it, just do it, and if anyone complains worry about it then?

Guus: yes, I saw that

I meant that some of the links in the e-mail threads there are dead and I couldn't get all the context.

I'm +0 on this change and I would like to request the two other directors to vote on this, too.

2. Trademark Policy Fees

I read the card on Trello. My question is: for which application is this a problem that we need to fix?

Is this about the fee for last week's approved application?

ralphm: no, it's about potential applications that are scared away by the Trademark

So there are no concrete cases about people not applying because of this fee?

ralphm: not as far as I know.

ralphm: obviously, there is no way to know about all the people who haven't applied because they got scared away.

MattJ, Martin do you see this as a problem we need to fix?

Ge0rG: to be honest, if I saw a fee that was inhibiting, I'd try to ask what's up with that

ralphm: I've heard from many people, mainly from the OSS movement, that Trademarks Are Evil and that Jabber Is Evil etc. and I want to streamline the process as far as possible.

It is not that we *need* the fee itself, but one of its (initial) functions was deterring frivolous applications, i think.

To be honest this still seems reasonable to me

We have effectively deterred most applications.

I'm not sure if people that think that Trademarks are evil and by extension the Jabber mark, will start using the name if the fee is $0

ralphm: I'm not sure about that either. I only have a slight hope.

It is still a trade mark

Yes :(

I have seen two out of three present directors not seeing a need to change this, so I have to decline the request for change at this time.

Martin?

ralphm: is your position that there is nothing we can do to revive the Jabber term?

not at all

I agree with you ralphm: it's still a trademark, and we can always approach/waive it on a per-case basis

Martin: the board has to decide about applications on a per-case basis anyway.

Ge0rG: I fact, I strongly believe it is possible to start using the Jabber mark for a marketing effort to attract people using XMPP-based chat. I do think, though, it would be good for such an effort to include the people managing the jabber.org service, and it also doesn't necessarily need to happen through the XSF.

ralphm: I believe that as a prerequisite, we need to simplify the trademark usage guidelines as far as possible, and that removing the fee is a useful step in that direction

I disagree

ralphm: with which parts of my statement?

all of it

in the interest of time, moving on

3. SPAM working group

Was this meant as a request to form a SIG?

Martin: I see you added this, so I assume you know. But I think what should happen is that somebody writes a XEP for this.

I'm not sure what the formal type of group is. I'm looking for a closed-membership group that can coordinate and devise a strategy before going public.

Dangerous assumption. Given most don't have access to add cards in Trello, I added it on request.

I'll note that there was some concern (I don't remember who from) about the group being closed

yeah, that bugs me too

I think discussing anti-spam techniques publicly is not necessarily a good idea

sure

The problem is that if they aren't open, they can't be implemented by operators

Right

Unless we form a small "special" federation of operators

it's just an arms race, discussing in private will at best delay them a bit, at least in public you can come up with things that can't be evaded

If that's going to happen, operators are free to do that, but I don't know if that's something that should/needs to involve the XSF resources

which doesn't require any XSF involvement (same as the operators list doesn't)

I'd like to have a small closed group for testing and strategy, that communicates and provides hints to other operators in a semi-public way

doesn't sound very federated

Ge0rG: right, so I don't think you need the XSF for doing that

ralphm: I hoped to get XSF blessing at least ;)

Ge0rG: like XMPP, the community is federated, we don't need the XSF to be at the nexus of it all

If this group is using XSF infrastructure (servers, mailing lists, etc.) I would personally like it to be public. I would want to follow along.

Ge0rG: you seem to assume this in all of the things you talk about, but I think a better approach is to find like-minded people and just start doing things

The operators list is a great place to start this

I agree with Ralph. And count me in as a like-minded person (who has some non-public anti-spam projects)

yay

I just don't think the XSF being involved in such a group is necessary or a good thing

So looking at the clock, I'm going to close this

But if that group ever needed something from the XSF, we can discuss that

4. Next Meeting

+1W?

+1W is ok with me

MattJ: indeed

5. Close

Thanks all, sorry about starting late.

Thanks very much for the feedback.

Thanks ralphm :)

I also forgot to ask someone to write minutes :-(

I can go over the log and send some out

MattJ: ☺

Tx for extended meeting

+1

MattJ: Can you explain what the deal is with https://xmpp.org/extensions/xep-0313.html#example-23 ?

iq-set returning stuff is pretty uncommon

Zash, I find it logical ✏

It returns the actually applied preferences, which may be different. that’s less racy and less RTT than requiring the client to query it again after setting

There's still the race of another resource changing stuff between your iq-get and -set

Zash, we have that race all over XMPP

MattJ: Rosters and blocking would like to have a word with you

You think there's no race there? What?

It'll tell you if something changes

I agree that a push method for preferences may be (have been?) nice

And those also deal in deltas, not the complete things.

But that doesn't solve the race in any way

What exact race?

You'll never know if another resource has a conflicting change in flight, or a push is already in-flight to you

MattJ: The mam-prefs fixes that?

No, I'm saying that it's no worse

And you'll notice eventually if there was a roster or blocking thing in flight.

You *could* fix that by having a reference to the current version (as you see it)

And then you can build a blockchain on that!

That's why I said push would potentially be nice

But it doesn't solve a real problem in any way

It's already an established pattern

There's little difference between push vs. pull

I agree, in hindsight maybe adopting the same model would have been better

It's a little more complex, but clients are used to it (well, they should be)

MattJ: Was that in the very earliest drafts? Or, how could I have missed it all this time?

I remember asking why it wasn't using the common push pattern at some point, but not having seen that it reflected the data back in iq-set.

afaik it's always been this way

It's almost as if the various xeps were written by different people :)

jonasw: found the thing that originally produced xeps.xml: https://q.zash.se/ef274a7e18fc.txt (grep for botsFile)

From the old all-in-one repo