XSF Discussion - 2014-02-19

lloyd: https://twitter.com/disruptivedean/status/436063951932379136 -- you should make him happy by showing webrtc in xmpp-ftw (-:

berling ^^is that our capitol's asian outpost?

FOSDEM 2005, and a beer event at Le Roy d'Espagne. Those were the days ;)

erm... is openfire seriously asking for ?username=null&password=null&resource=xmpp for websocket connections? i've read the source but...

Eh, wat

you specify as ws://yourserverhere:7070/ws/server?user=null&password=null&resource=foo uri

and the source has if (username.equals("null") && password.equals("null")) // anonymous user

Has anyone done any thinking about storing read/unread message states or read-up-to-<messageID>? Would this be something added to MAM?

Kev - I think you mentioned having done some thinking about this.

I think there was some talk about this on the mailing list recently...

I'll check the archives. I know there was talk of needing this during the Summit.

Lance and I discussed this outside the lift in the Aloft.

pity you were only on the first floor Kev.

Kev: is this an addition to mam or new spec?

Anyway. I think there's a little jiggling to go on.

Is that a technical term?

In as much as what happens is that you use carbons + the MAM id (which answers my question about whether it's useful) + the 'reset' that we discussed in terms of the push spec.

The reset gets relayed to all your clients because of carbons, and boom. Everyone knows everything's read.

Doesn't that makes the assumption that one is using push?

for example, it would be nice to have a webclient that, when you log in again, shows you an unread counter for posts newer than <last time you logged in>.

simon: It makes the assumption that we define the 'reset' outside the push framework.

Kev: if I understand you correctly it's a "I've read all messages from this user/pub-sub node/mam-able feed". I thik we're looking for something a little more granular (but not called IMAP).

as in I've read up to this point.

I'll write up the problem to the list if that makes it clearer.

It's I've read everything up to point X in my stream (and therefore in your streams too, because of carbons, as far as chat messages are concerned).

that sound exactly like what we want for each channel: A client can push back a read-to-this-point to the server. Next time they log in, the client can show you that you have n unread posts. What's the next step for speccing this?

simon: I once thought up a way to implement notifications like on android devices. Every notification is a pubsub item. You can retract them. Done.

ralphm: true. I'm pleased that we discussed the need to retract push requests at the Summit - something that could have been overlooked easily.

I think I like how it works. Pushes yield notifications and retractions do as well.

ICMP_ECHO_REQUEST

ICMP_ECHO_REPLY

RESPONSE, I thought. But I was hoping for some fellow Board people to respond. :-)

0x0303

Zash, yeah, I can't even be bothered to look up the ICMP type.

So I think it's Board Time. Isn't it?

Hi all

ralphm, Ping?

Testing testing

Laura, Works.

Thanks! Had issues connecting

We'll give it 5 to see if ralphm or bear show up.

cool

here

ralphm, I'm assuming you'll act as Chairman?

Wasn't expecting my lunch to arrive at 11:30. :/

Sure

0. Opening

Welcome (almost) all!

Agenda items?

1. AOB

simon: don't do that

No other agenda items.

GSoC?

Kev: ok

Ah yes.

ralphm: Now?

Kev: not yet

simon, Laura, anything to report re. website?

(yes or no is fine at this point)

dwd: any items you'd like to add?

No updates as yet. Planning on having something to show next week (not images but the outline plan)

ralphm: unfortunatley not.

Sorry

that's fine

I don't think I've much.

ok

FOSDEM finances, I suppose.

laura: we really need to get back on that website horse.

Simon: I will drop you a message

1. GSOC

This is supershort.

Kev: you have a report?

The XSF didn't apply for GSoC.

At the time of the application deadline, none of the ideas on the page were complete enough.

That is all.

Kev, Thanks for your efforts here.

Were you disappointed?

i've heard jitsi applied... not sure why they didn't try it under the XSF umbrella for related stuff (@emcho)

Yeah, I think GSoC's a great thing to do. But I don't have time this year to commit to mentoring any Swift projects, so I can't point much finger of blame at anyone else for not having time either.

Maybe one for us to bring to the agenda earlier next yeat? More time to prepare?

And while we could have tried our luck with the page as-is, both Bear and I felt that putting in a sub-par submission was the Wrong Thing to do.

Agree

fippo: well, jitsi usually applies for themselves

same for FOSDEM

also, XMPP is not all they do

Laura: It needs someone to act as chief cat-herder to get the project ideas in shape. In previous years I've done this, and I think when Bear volunteered to be org admin, he hadn't quite twigged the amount of effort it takes to cat-herd behind the scenes.

Kev: thanks for your efforts, anyway

ralphm: they're coming closer (-:

fippo: but hey, maybe you can heard projects for 2015?

Got it

So it ended up being a case of seeing if projects would get their ideas in gear themselves, and that didn't happen.

But yes, starting earlier next year, and coming up with some 'Your ideas must look like this as a minimum' would help.

Anyway, I think I'm done for this item.

Thanks.

Kev, Could the XSF make this easier for next year, by ensuring there's a communications channel to open source project leads or something?

Maybe a template or something.

To help identify what we need to support it through

dwd: Yes, this has been me in recent previous years.

Kev, Right, which is not ideal, especially for you.

That is - the org admin (or someone similar) needs to essentially chase down every project that might contribute ideas, and arm-twist them to do so (or not participate), and then chase them down to make sure all the ideas are in order.

I actually like doing the org admin role, I think GSoC's a great thing to support, but ... not this year.

Right, I was wondering if it'd be easier if all the project leads were on a single mailing list or something.

In my experience, it takes direct 1:1 chasing.

OK.

Not that people need chasing because they don't want to participate, but people need reminding. A lot :)

Kev: what would be a good time to start for 2015?

I think starting in November/December is sensible. If we had a cheat-sheet for writing project ideas (I'm sure we've had this in previous years) available by the end of the year, that should be enough time for people.

Do you think it would be good to have some examples of successful project applications so that projects know how much detail they need to go into?

Assuming GSoC2015 happens, which is never a given until it's announced, etc. etc.

That is always a winner

Makes it seem less dautning if you know what 'good' looks like

simon: I think we've had a template in previous years, yes.

Although we learn each year a little bit more about what's good - both from our own, and from the chats that we (by which I mean I) have with the other orgs participating.

Template is one thing, but pointing at a successful project might also be good. Anyway. I think we're all a little overloaded this year and have more than enough to be getting on with, without GSOC too.

Rigt.

Right.

Kev: I'd be very surprised if there wouldn't be one in 2015

ralphm: As would I. But it's not a given.

Sure, but would time spent on it be wasted?

Yes.

But I think it's likely enough to go ahead anyway.

Kev, Were there are project submissions that seemed particularly worthwhile?

dwd: Yes, the BC and IoT ones - they just lacked some of the bits they needed.

OK. I'm wondering if these things would generally benefit the XSF and the XMPP community sufficiently to warrant us putting some money that way. But this is just a thought to float about rather than something we should debate right now.

I think we could agendaise that for another time when there's longer to discuss :)

Right.

all right then

2. FOSDEM/Summit finances

Just a note that due to travelling and generally being rubbish, I've not sorted out invoicing, and won't have managed to before next week at least.

ralphm, If you want to accelerate *your* reimbursement, I can give you the revelant details to have Isode pay you direct as we discussed.

dwd: sure. I have to scan my receipts, too

ralphm, You'd just need to write an invoice. I'm not going to get a chance before Tuesday at the vert earliest.

dwd: that's fine

3. AOB

I wanted to ask about our communications in general

Once upon a time, we had a Communication Team.

Quick reminder to fire up your TLS engines running on Saturday for the second Security Test Day.

It seems at least some of the functions are done by Simon and Laura now.

But it is unclear to me who is responsible for our "social networking" outlets.

ralphm: yes. I'd like to do more.

Bear has registered this twitter account last week

ralphm, Yes, this is true. I think it's important to recognise that we have both internal and external communications to consider.

ralphm: which one?

but Neustradamus, who used to be on that team, has access to facebook pages and what not

Agreed. I would really like to know a) history of comms and b) what is happening now?

apparently he has communicated this with (former) board members at some point

Neustradamus is very keen to help btw.

http://xmpp.org/participate/become-a-member/the-xsf-communication-team/

This is our description of its functions.

Laura, What's happening now is more or less whatever you and simon are doing.

So not much then.

I want to know more for the website work I am doing

More engaging, less secret chats etc

I am not interested in the volume of of communications for this discussion, but rather it being a concerted effort

Let me put some thoughts together and share with Board?

For initial discussion

And, if you need help, you may wish to fire up the above team at some point.

Laura: sounds good.

Leave it with me. Will have this with you on Friday/

Laura: cool. Be sure to check out that link I pasted above.

Laura, I'd be perfectly happy if you want to just take it over, to be honest.

hi bear

Will do, and Dave my thinking is along those lines

We're almost done here.

Laura, What I know about proper internal/external communications could be written on a postage stamp. I think you're the domain expert here.

hello - sorry for being late

bear: do you have anything you'd like to discuss?

I can't contribute much technically, but marketing is what I do

It is my job so I should be!

no, just to add that i'm still working on getting a list of tasks that were deferred

Lance: perfect

Laura: perfect

(but you are great, too Lance)

bear: ack

Any other any other business?

4. Time of Next +1W 5. Close Thanks, all!

Bye!

laura, dwd - if you want to discuss GSoC and how it (doesn't) work I can do that at any time

bear: what about half an hour ago?

half hour ago I was outside tryiing to keep the blocked storm drain from flooding my basement :/

Oh thanks Bear!

bear: but you said "any time"

I just wanted to point out that a lot of what they mentioned had been done - but it's like what Kev said, it requires a lot of cat herding

bear: I'm sorry. Flooding sucks.

no worries, I knew you were joking about "any time"

i'm just cold, wet and flustered at almost everything right now

bear: after initial refusal, some Dutch engineers have finally been taken up on their offer to help out with the floods in SW England

Oh, hey. I realised we've not discussed the IETF thing at all.

they came in with 30 trucks of pumps

you think the UK would gladly accept Dutch engineers given their experience with keeping whole cities from flooding

ralphm, FWIW, there's been artificial drainage there for over a thousand years.

and had to stop pumping, because they pumped too fast

ralphm, Most of the rivers there are actually artifical drainage channels.

So this IETF thing - where was that Wiki page bear did?

dwd: yes, but it seems kinda silly to refuse help for weeks

http://wiki.xmpp.org/web/IETF_89

So this meetup thing

bear: I suspect "The UK" as a whole would gladly do so.

I don't think it was exactly a country-wide decision not to have help with the flooding.

(Or, indeed, that it was generally known that help had been offered and refused, until much later)

ah - US centric thinking - for us it's often the federal government who handled emergency items like that

bear, So my take is that all those people would be funded by their employers, most likely.

bear: Yes, it was the central government that refused the help, as I understand it. Just that I suspect this wasn't a decision that would have been supported by much of the population.

yes, so far everyone who has ssigned up seem to be corp backed

bear, Central Government, yes. The whole of the UK, no. Also, I doubt the dutch engineers where offering for free. :-)

Kev: from what I heard, Cameron himself had refused

As there are people going who have never attended IETF stuff before, and given people need catherding, as previously mentioned, it might be helpful for someone to send out a mail to the list with what needs doing for those attending.

dwd: I wonder how costs weigh up against damages

ralphm: Yes, I'm not sure if you're implying that as evidence For it being widely supported or Against it.

dwd: No, I imagine not.

kev - i did send a list

Ahhar.

err an email to the list

Kev: I haven't looked for a credible source

(Channel 4 isn't, as far as I know)

ralphm: I was implying that Cameron and the Will Of The People might not always be completely in step :)

might always not be <= fixed that for you

bear: If you meant your "IETF 89 and the XMPP WG meeting - are you going?" mail, I was thinking of something a little more prescriptive

ah

But I seem to be the only person attending who probably hasn't been before, looking at the list on the wiki, so ignore.

I thought there might be a wider audience.

so did I

the push back last week, in my understanding of it, was to avoid actively stating that people who wanted to attend could get daypass covered by us

so I avoided any direct wordage

Right, which, I suspect, means that a lot of people will have been put off.

That said, xnyhps - you're coming, right?

Yes. I was just thinking that it's about time I'd get myself a wiki account. :)

The other WGs that are relevant to us are UTA and DANE, both (I think) on Friday.

It'd be very useful to get Zash there, IMHO. He's done more implementation of this kind of thing than anyone else in the community, I think.

.

xnyhps, Are you there for the whole week as well?

No, I'm leaving on wednesday.

Shame to miss the DANE meeting, but to stay two extra days just to wait for that...

xnyhps, UTA as well, which Alexey pointed out to me.

True

To my shame, I hadn't noticed UTA. :-)

I'm assuming perpass is also signficant to us as a community; I've no clue at all when that's meeting.

UTA?

Kev, Alexey said it was something to do with TLS usage in applications protocols, most specifically email and XMPP.

Kev, Think RFC 6125, basically.

Ah.

calendaring fail

will read the logs

http://tools.ietf.org/wg/uta/

bbiab

Oh! ISOC members seem to get into the IETF on Tuesday 4th March for free.

and being an ISOC member is a good thing anyway

which reminds me...

dwd: really? I hadn't known that - but it's easy to become an ISOC member :-)

stpeter, Right, I am already. I think this may only be "local chapters". Which I'm technically not in, though I'm in the England one for convenience.

yes, I'm in my local chapter as well (Colorado)

Might have to do that one. I qualify for both a WG Chair dot *and* a personal mentor to show me aroound, which is fun.

$19b. much.

for a broken xmpp implementation and a gazillion users

... that already are on facebook using XMPP

… for which they now have a very nice network graph based on phone numbers.

yeah...those centralized folks have it easy in that regard :) don't have to care about privacy an all

must be living the dream

I don't think they didn't already have that

Tobias: yeah

:|

They now have a lot more phone numbers. And can probably correlate numbers based on the graphs for users that stubbornly didn't provide their numbers to Facebook yet

... like me

… and me.

intosi: i.e. 99% of fb users has already uploaded their addressbook

That's unfortunately very much true.

so, they gained. um. groupchat?

FB already had that.

not as persistent rooms

Close enough that end-users don't see the difference.

At least, 99% of all end-users.

no, this is actually hard in facebook's app

ralphm: Not sure if WhatsApp is only XMPP without federation, it should partially contain HTTP also.