XSF Discussion - 2019-01-10

pep.: you're not an issue :)

It was more a general observation that we need to control access somewhere

Guus et al., going to fix my council bio in the afternoon

not sure what to put in it, never worked with XMPP professionally :)

(unless you count the failed experiment to save XMPP from being ditched in the company I worked for at that time)

I thoroughly disagree that we're controlling access in any meaningful way.

we're making it hard for people to be effective, yes.

for the xmpp.net thing, I think we are

or at least, we can be, if iteam reviews the diff before using a new image ;-)

but there's hardly an access policy being enforced.

@jonas', in your bio, quote your role as editor, and one or two contributions you made over the year. That'll quickly get you to a couple of lines of text.

is your client doing the '@' thing or are you?

I am, in an attempt to make my client do mentions. It triggers autocomplete, but I think a bit of parsing fails properly recognizing the ' being part of your name

so, it it can't do mention, it'll send the @ symbol

jonas’ <-tab completed

@jonas' <- typed manually

right,there's a difference there.

the tab completed one does work as a highlight in my client, the other one doesn’t :)

fun bugs to discover :)

Yeah, the mention is in context of xep-0372

I found another bug where it's case sensitive too

and, as the failed mention neglects to remove the @ symbol, some clients fail to properly recognize the name, resulting in no kind of notification being triggered at all

quite the opposite of what was intended

took a while to figure out why this was happening 🙂

https://github.com/conversejs/converse.js/issues/1401

Guus, the reason why it doesn’t tabcomplete is that ' (what you are typing) and ’ (what I am using) are different characters :)

ah. My font renders them the same

it’s a trap :)

dwd: yes all good now (sorry I wasn't paying attention to the chat last evening)

updated / closed the ticket

Upon (very) close inspection, my font _does_ render them different. I think thre's a one pixel difference

Guus, lol

you could still argue that converse should normalise those by the way, because it’s hard for non-weird people to type those

although the NFKC procedure doesn’t normalize it, so it would have to be a custom lookup table

meh. I'm shooting to much issues for that project as it is 🙂

could you / someone confirm that this is roughly what we want, summit announcement wise? https://github.com/xsf/xmpp.org/pull/503

I think there’s a more proper way to link to the page from the markdown

I’ll have to look into it, one sec

yeah, try {filename}pages/community/... instead of /community

that’ll use pelicans linking mechanism which is more robust against use of save_as in the page itself

Guus: the content looks good

@jonas' I dont' understand

hargh

jonas’ I don't understand.

substitute: * [XMPP Summit 23, Brussels, Belgium](/community/events/summit23) with * [XMPP Summit 23, Brussels, Belgium]({filename}pages/community/events/summit23)

jonas’ will that work in that index.html too?

I"m not sure if that's parsed?

I’m not sure about index.html either, so let’s play it safe there

(and I don’t have a xmpp.org checkout at hand to test)

change pushed.

jonas’ like that?

oh, that actually doesn't work

my docker instance generates a link to http://localhost:8080/community/%7Bfilename%7Dpages/community/events/summit23

meh

then revert

sorry for the misinformation (maybe it has to do with our pelican version?) but actually that’s supposed to work

no problem

merged

\o/

So, I was annoyed by the manual copy of sidebar menus, and did this to "fix" things: https://github.com/xsf/xmpp.org/pull/504

reviews are very welcome.

goffi, Thanks for the pointer.

Guus, request a review from me, will see if I can look into it this afternoon

jonas’ done

MattJ: https://github.com/xsf/xeps/pull/696#discussion_r213904512 stumbled upon this

This is wrong on so many levels, I can't even...

Replied, thanks :)

Someone from the Berlin XMPP meetup likes to write on the wiki. It's about the sprint. Who should be contacted to create an account?

debacle: I can do that

needs a username in wiki CamelCase syntax and an email address for password delivery

I pass your JID to him, Ge0rG?

debacle: feel free

thnx

Guus, MattJ, my dearest congratulations for xmppoke/xmppoke-frontend and XMPP.net!

Thanks!

I'll be 5-10 min late for the meeting

sorry

Ok

XSF Board Meeting | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

0. Welcome + Agenda

Hey

Hi!

hello

Who has anything to add to the agenda?

Nothing here

Agenda has been filled over the last few weeks, on trello

(I'm hoping you guys try to keep track of that during the week, in order to prepare)

I do, but I still ask anyway.

1. Minute taker

👍

who?

...

2. Commitments

Sending out an e-mail to finally schedule a meeting with Peter today.

3. GSoC

here

Guus, I suppose no admins have signed up?

Actually, one has

Florian / flow volunteers.

oh, cool

To quote him:

> I would be happy to help the XSF and the XMPP community with GSoC, including acting as an organization admin. > I've been a mentor three times, visited different GSoC events (mentor summit, local meetings) and know various other GSoC mentors and admins from other organizations.

That's great!

hargh, doorbell, one sec

and back

Sounds awesome, flow thanks

I suggest that we let flow take the reigns on this, and see if he can organize an XSF umbrella for GSoC involvement of projects that are interested.

I think that Kev was our last admin, so I hope he can answer any questions Flow might have.

Hmm?

Oh, GSoC.

yes

I motion that we bestow upon Florian the role of XSF GSoC'19 admin, and that we ask him in that capacity to organize an XSF umbrella for projects that which to participate.

We must have co-admins BTW.

+1 and +1

+1

So Flow on his own isn't enough here.

Kev: to review applications, you mean?

(At least in name, even if they don't do anything, there must be other people willing to do stuff if Flow doesn't)

No, I mean Google require there not to be only a single admin.

How did we do this before?

I found people who trusted me enough as admin to put their name down in the expectation they didn't have to do anything, generally.

Arranging for a correct registration with Google shall be up to him, then.

We wanted someone to lead the effort - I feel that we now have one. He can take it from here.

Feels like a solid start +1

+1. As it is a requirement to complete the application, I am curious to see who Flow will draft.

I'll let him know that we accepted his offer to act as admin

3. FOSDEM / Summit

registrations are pouring in.

The FOSDEM organisation has a sent a provisional overview of the stand assignments. I am a bit worried about it, as it shows 5 tables in the same area as before, but just one for us..

(For GSoC - I'd like to note that before the application we need to have our ideas page in order, etc., so we should make another call as we approach that date as to whether to actually apply, based on the quality of what we've got. We shouldn't burn the goodwill we've built up over the years with the GSoC team)

So I am going to try and see what we can do about this.

ralphm that'd be good, thanks

Kev: noted and agreed.

I'm waiting for a sign-up form from the hotel (they appear to be having difficulties generating one, but did already receive my signature on their offer.

There's no alternative than wait?

I am almost in contact with the regular restaurant for the XSF Dinner (we keep missing eachother), and we have had a sponsor for said dinner step forward.

I pinged them, they told me to wait longer.

Ok thanks.

Also yay for the dinner sponsor.

one thing that concerns me is a lack of salesware

if we do want to sell stuff, we should design/order things

Yeah, I was thinking about this too. We have no swag left, in terms of clothing.

Last year, there was a suggestion for a Summit-only t-shirt. I actually made a 'design' for it.

we're 10th Jan, it's late...

I mean to generate clothing

I disagree

lets take this to the SCAM muc after the meeting

yeah

4. Pseudonanymous Membership Applications

The annual discussion? :)

I see there's a question on whether members can be anonymous or semi-anonymous. I think this was addressed in earlier years and the answer has always been no.

Is that an anuual thing? Never heard it before.

Me neither.

How did this come up

People have too much time.

ralphm: I was asked whether it's possible.

SOmeone asked Ge0rG , who asked us.

All I found was the XEP that's next on the agenda

I am happy with not allowing it.

+0

Yes, it's been discussed extensively in the past, and there are reasons it's not feasible or sensible (as nice as it may be in theory)

There's no vote on this. I think it comes from our being a registered corporation for the state of Delaware.

There appears to be precedent to not allow it. On top of that, there might be legal issues related to accepting anonymous members into an organization where they can hold legal status.

well, ok then: no.

Guus, exactly

I'm happy with whatever the official answer is, and even more happy if it gets documented.

Ge0rG did just volunteer to do our minutes, I think. 🙂

Heh

Ge0rG, thank you

AS Ge0rG suggested, https://xmpp.org/extensions/xep-0345.html might be a good way to formalize this further.

You don't want me to write minutes in my current mood, sorry.

I'd be much more grumpy than usual. Also I'm affected by today's decisions.

it has been deferred, but has received earlier updates based on feedback. It has some editorial issues (spelling/punctuation) but looks good otherwise.

I'm not opposed to resurrecting that XEP.

I'm not opposed either, though I still have some issues with it

Is the other item on typos referring to this XEP specifically?

The XEP contains a clear statement that members must apply with their legal name.

ralphm I dont think so

I think the typo thing is a general issue

ok

Shall we ask the Editor to LC this XEP?

Then we can collect feedback and resolve issues we have with it ourselves.

I *assume* that the typo thing is about whether a Deferred XEP should be re-promoted to Experimental when somebody fixes typos in it.

I found a XEP of mine that should be deferred but only isn't because someone fixed something editorial

deferred -> last call is a state transiation that's allowable? ✏

MattJ if you *want* a XEP to be 'deferred' maybe you should retract it.

I don't, I'm glad it's not deferred :)

But it still feels wrong

there is another processual thing coming up, regarding https://github.com/xsf/xeps/pull/727 - we can't move from "Deferred" to "Deprecated" currently, and that might be useful.

ok, to many topics at the same time now, I think 🙂 I'm starting to loose track.

(Although I did ask that to be run past the community, since Council isn't in consensus about that)

What dwd said

On XEP-0345.

We can ask the Editor to undefer it.

Then put it to LC. I'm not sure if that requires any changes to the document.

I'd be happy with that

XEP 1 talks about updated versions of XEPs, and I don't think a typo can be considered an updated XEP in any substantive way, it's an editorial change. Until recently we wouldn't even have updated the patch number.

Yes, but for XEP-0345 specifically, XEP-0001 says the Editor can decide to change back to Deferred at will.

(This agenda item is *only* about members applications, really)

you could vote on asking the Editor for doing whatever is needed to LC 345

i put details on the typo thing in the card description

So, besides Guus, who else agrees we will ask the Editor to put XEP-0345 to LC?

I agree

nyco, Seve?

regarding deferred -> lc, I think theres something in the prose of '0001

(thanks ralphm )

jonas’: I read it, not directly, but I think it can be done as described above.

jonas’ We'd like to revive the XEP, with eventually it going to LC.

sorry I'm on mobile

I agree, ralphm

> Note that if a XEP is Deferred, the XMPP Extensions Editor may at some point re-assign it to Experimental status

if deferred -> LC is not a valid transition, then we can go through one that is valid first.

+1

Let the records show that the Board would like the Editor to move XEP-0345 back to Experimental, and then, in a separate step, put it to Proposed.

5. E-mail issue

I understand the Seve has now received Board e-mails.

emails from wiki work again, seve receives emails too now.

yay!

seems resolved as far as I can tell

Yes, very happy :)

iteam let me know that action was taken to remove us from a blacklist

6. GitHub Issue 497

https://github.com/xsf/xmpp.org/issues/497

does an LC for '345 go to standards@ and/or members@ or what? it's procedural and affects membership

I put that up, before the comments on that issue poured in

jonas’ when in doubt, spam everyone.

Guus: I understand

jonas’: to members

ugh I lag a lot I think

As for 497: I think Peter implicitly acknowledged that such an agreement was once made. Therefor, I feel we should honor it and restore the thingy on the website.

This is about the xmpp.org domain having been donated, and requiring a notice to that end on our website.

We should clean up our list of sponsors, and just put a logo for OpenDomain in that sponsor cloud

we'll work with them on the 'how'

my primary concern for putting this on our board was to see if it was legit.

that since has been established.

Yes

ok

7. Clarify typos in Deferred XEPs

jonas’ wrote: > I (Editor) ask that we clarify XEP-0001 in such a way that editorial changes to Deferred XEPs do not automatically un-defer them, while non-editorial changes do.

that seems sensible to me.

I think it's already that, pretty much.

I don't think that fixing a typo is considered a change if it doesn't change anything substantial.

I'm happy for the Editor to make that call.

XEP1 talks about x.y versions as being updates, and we publish editorial under x.y.z.

we might go one step further and state that editorial changes do not change the state of a XEP at all. ✏

Tightening up the wording here seems fine, but I don't think this is a change from what XEP1 is telling us to do.

+1 Guus

even better

Yep

So we don't have to do anything?

just confirm to our editor that he's right.

well, he asks us to clarify the text

we'll ask him to draft a proposal for that clarification then 🙂

Ok

8. AOB?

nope

(some of the other items are pending as per agenda item 2 above)

i don't have an AOB.

Nothing here (I do have some low priority things but we're already quite over time and I have another meeting in a moment)

ralphm: do you have an estimate on how long the resolution of the JabberSPAM trademark request will take?

Ge0rG: not yet

9. Date of Next

We are almost a month in :)

+1W

+1

@Mattj maybe add them to trello for nexzt time?

Yeah

10. Close

Thanks all!

thx

Thank you

XSF Meeting | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

Thank you guys!!

Thanks all

different topic now, I wanted to know if some of you were interested by some 2FA mechanism for XMPP ? for example by adding an extra step to SASL ?

edhelas: IIRC there was a proposal last year about that. And there are also interesting ideas around per-device passwords and tokens and the like

okay

edhelas: https://xmpp.org/extensions/xep-0388.html

XSF Discussion | Logs: http://logs.xmpp.org/xsf/ | Agenda https://trello.com/b/Dn6IQOu0/board-meetings

Ge0rG When I asked Daniel about per-device passwords or Web-API-style tokens, he mentioned also client-side certificates. Not the same thing as 2FA, but related issue.

Mmmm https://xmpp.org/extensions/xep-0257.html

I seem to recall that dwd worked on this.

ah, he's the author of XEP-0388, so yeah.

I updated/archived a bunch of Trello cards.

Thanks

jonas’ do you need further info on either question you initiated?

Guus, the typo thing seems to be well-addressed to me

The LC thing is mostly clear, but I’m not sure why you’re insisting on having this in two separate actions?

you asked for a change in XEP-0001 - can you prepare a concrete proposal?

Do you want two separate emails?

Guus, I can, can you send me an email about that?

jonas’ please use the optimal path that's available. I'm unsure of a state transition from 'deferred' to 'last call' is permitted in XEP-0001.

we simply want to revive that XEP.

For those no in summit@muc.xmpp.org, yet, but interested in talking about all things FOSDEM / Summit, please go there.

(no->not)

jonas’ mail regarding https://xmpp.org/extensions/xep-0345.html sent (to members - with an invitation for anyone to review)

edhelas, Yeah, At Surevine we built a web-based 2FA thing around SASL2, TOTP, etc. The XEPs and one internet-draft came out of it, as well as some code (for Openfire, open source, rough shape).

edhelas, Surevine still have a stanza.io implementation, but I don't believe that is or will be open sourced.

Half-Shot, hey, do you have any estimates (or resources showing these estimates) on how many matrix servers there are out there federating?

ah I'm interested

dwd what kind of 2FA did you used ? the Google Auth thing ?

edhelas, https://kitten.ietf.narkive.com/OsJEMivx/totp-in-sasl-in-xmpp is a breif for the Kittin folks in IETF. But yes, we used TOTP.

pep.: lol

oops

edhelas, Also: https://mail.jabber.org/pipermail/standards/2017-August/033148.html

> draft-cridland-kitten-clientkey :D

KITTEN

edhelas, The IETF used to have a working group called CAT, for Common Authentication Technologies. When it was restarted later, it was obviously called KITTEN, as "Son of CAT".

someday we'll write books about that

Return of the Two Cats

edhelas, Anyway, it all worked with Google Authenticator and stuff. Handled "remember this browser" stuff and so on, as well as handling session reconnect without re-TOTP (or manual reauth, sometimes).

Guus, I’m not sure whether "such a transition does not exist" is meaningful when the editor can Un-Defer at will

Guus et al.: The normal Last Call template (see for example https://mail.jabber.org/pipermail/standards/2019-January/035623.html ) does not really fit the XEP-0345 last call.

what do you suggest as "guiding questions"?

Does this do what it says it's doing?

Does this even make sense?

Would you implement this in your wetware?