XSF Discussion - 2013-09-18

hi,guys

anybody here

oh,lazy guys

why not getting up

weizfan_1111: how can we help you?

i tired of silent

why not say something

join another room, this is the room of the XSF here, if the XSF has nothing to discuss its silent

okay

Alex,can you recommend a c++ library for server?

jdev@conference.jabber.org is the venue to ask this question

its all about xmpp development, many developers hang out there

hey all — is it board meeting time?

yes

With an audience no less

the council meeting is still going, so we may have a couple of minutes before starting

oh weird

I don't see bear here

might need to rejoin

he's a moderator

bear is in the list?

s/?/.

ah, that's better

presence issues

anyway, joined from another account

swift has been going into a zombie mode when my laptop hibernates

It probably doesn't notice until TCP times out. We should fix that.

Ashley, Arc and myself - present

Florian?

Jason?

i see Florian in the room roster

Jason didn't reply

Hi all

simple majority for a quorum

hi Florian :-)

well we certainly have quorum now

4 out of 5 - let's call that a win and start

drumroll

remember to update your bios for the 2014 board election folks

was the primary agenda item engagement with that other standards body? iirc

yes, that is the reason for today's meeting - for Peter to give us the info on what they want and what resources will be required

right

so, we received a "liaison request" from a technical committee at ISO

TC 122, which does work on logistics and such

it's sort-of "Internet of Things" related

now, the ISO is a lot more formal than the XSF :-)

:)

however, essentially they would like some folks to review their work on a usage of XMPP for notifications related to things like package delivery and vehicle tracking and such

unfortunately, regular folks can't just join those discussions

because of their access controls over things they are working on

so they would like to establish a liaison relationship, which would enable us to assign a few people to participate in their work and sanity-check what they're developing before it gets released

is the liason allowed to send the info to others for review?

as I understand it, we would assign two or three people to review / participate

there are a few questions that arise, of course

e.g., how do we assign people (ask for volunteers among the membership)?

(I can mention that myself and Joachim have worked with them for larger part of the year)

since it's IoT-related, some people here might be interested

(and that original participation request was sent on the IoT mailing list about a year ago)

Peter Waher: ah, are you guys participating already through normal ISO processes?

yes, through what is called IEEE/IEC/ISO P1451

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

this is related

another question is whether people we assign to be liaison experts are taken as speaking for the XSF, in which case we might want to have a consensus process for formulating our feedback

they are various groups looking at similar thing

s

Peter Waher: yes, IoT stuff is hot

how we handle these things on our side is up to us — e.g., we could assign just one person if we so please

I'm president of a subgroup, relating to xml & xmpp telegrams

or we could tell them that they already have some folks involved who know about XMPP and thus we don't see the need for a formal liaison relationship

do we anticipate changes to XMPP itself, or simply new XEPs?

that's why I was asking if the liason is allowed by the rules to pass on information for review

Ashley: not changes to XMPP, and probably not any XEPs

ok

during the work we've done together, we've proposed XEP 0322-0326

Peter Waher might know better, but I think they would write a document that says "here's how we're using XMPP for our use cases"

This, they have done

I see it as similar in some ways to what the OpenADR folks did with OpenADR 2.0

and probably want us to double check

right, make sense

Peter Waher: yes, that seems reasonable

right

they have issues with legacy binary encoding of information and tagging

e.g., I completed a review of the OpenADR 2.0 text about XMPP and suggested a few minor fixes, but mostly it seemed reasonable and I told them so

so they probably want a third party to double check before publishing the standard

yes

thus one question is, do we see the need for a liaison relationship?

if the Board would like, I'd be happy to be a liaison just to do what Peter suggests about reviewing what they come up with

unlike OpenADR, ISO is more formal and thus requires a bit more process

A formal liason relationship would lend us some legitimacy, I'd imagine.

that seems like a good way to go

I'd say: why not... don't really see any downsides here

Dave: right

Which obviously costs us in terms of effort.

Dave Cridland: right

stpeter are you volunteering?

it helps to legitimize the XSF, perhaps :-)

arcriley: yes, I'd volunteer to be a liaison since all it really means is I review yet another standards spec in my life :-)

i look at it like building our ecosystem

Yeah

it would be great to build a broader concensus behind any IoT solutions the XSF proposes

Peter Waher: agreed

and spread knowledge of the area

Peter Waher: unfortunately we don't have a lot of participants who know that space very well

but that's a separate issue :-)

I'm available for any comments, questions or doubts you may have

And I'm sure Joachim is as well

Presumably that's what they're discussing, though, and what *we're* needed for is the XMPP side

yes

and we do have some amount of knowledge about XMPP :)

I'd personally suggest that Peter Waher, if he's willing, be the actual liason, and have the review done by a smallish team (and/or the Council)

it seems to me that they want a sanity check

In which case we'd probably want to pick someone well versed in the XMPP side of things.

(e.g. stpeter)

Yeah

Dave: well, Peter and Joachim already participating, it seems

^ + are

Peter Waher: are you able to participate fully already?

I can be the liason, but for them to be satisfied, it would be excellent if stpeter or the council could approve by double checking any recommendations

heck, I could probably participate with my Cisco hat on

yes

Right, but I'd expect the liason to be in the awkward position of acting as translator more than anything else.

Dave: translator how?

Translating both XMPPisms and XSFisms (and ISOisms)

It should be remembered that this request is probably originating in the desire to double check what has been proposed by myself and Joachim

in a more formal manner

Ah. That might be awkward. :-)

Ah, right

so, If somebody well established in the XMPP community (i.e. stpeter) or the entire XFS council would review any suggestions, it would address their concerns

"Yes, we'll check Peter Waher's work. We suggest Peter Waher should check it."

hehehe

:)

:)

Would anyone from the council be interested?

Peter Waher: it might not be that they have any concerns with the feedback you've provided, but they don't know XMPP and they want to be Sure™

Alongside stpeter

this is my sense of things, anyway

I believe so too

however, there

's one item of concern

Florian2: I guess I could.

Actually, we may not even need 2 people from the XSF

and that's legacy binary encodings

hey, we have binary XMPP, no problem :P

And I've been quite stubborn in explaining why it's not a good idea to continue down that road

stpeter: lol

for interoperability's sake

I've proposed the use of bits of binary, for instance, but it would create a solution nobody would be interested in

but it's not really for us to tell them what bindings they need to support, right?

exactly

i've tries to explain pros and cons for different solutions

and what I personally believe is the correct solution

now, there are 3 proposed solutions

it seems that their decisions about old binary vs. new XML (aka "legacy JSON") are something that's up to them

1) The XEPS 0322-0326

2) Bits of binary (not recommended)

(although we might suggest that they might not want to send their binary encoding over XMPP because it makes more sense to send the XML representation)

3) a combination, where they use telegrams in (1), but encode binary field names using some kind of urn scheme

ah

well, this is something to work out in conversation with them

I guess, if somebody could validate the pros and cons, they could feel safer taking a decision

I'd prefer not to talk about the specifics much here because it's their IPR and their processes aren't all that open

(y)

(which itself raises an issue about broader review)

They've sent us all documents, but we cannot forward them

But they have no problems sending them to any participants

so... the liason cannot forward them to the council for review?

but if we can assign a few people or just me or whoever to advise them, then personally I don't see the need to open up the review to all XSF members or whatever (although perhaps we'd see advice from the Council somehow if we see the need)

arcriley: good question

arcriley: I just thought about that 10 minutes ago and I do not know the answer, but I can find out

we've been clear on that point too: Any XEPs or documents leaving the XSF will be in the public domain

if thats the case why don't we just appoint the entire council - that way they can discuss and review while still complying with their closed process

They have discussed the possibility in publishing the documents in an open forum too...

Council discussions are (or have been, historically) public.

arcriley: that's a possibility, although Council membership changes and as I understand it they'd want people to be appointed as individuals

but today, the documents are still not public. Only they can send them to any participants

While it's within the Board's remit, I suspect, to propose the Council could hold those in camera, I'm not sure it's a precedent I'd like the Board to be setting.

right

I've been very consistent regarding openness

ISO has got to have some precident in working with organizations like ours

Yeah

They know it and respect it

So I believe we can push that point

this would have to be not a Council issue, even if all of the Council are part of the liason team

well, the easiest thing is to appoint one or two people as individuals and not say that the XMPP Council will be reviewing things as a body

that's what I was trying to say

because I agree with Dave about not wanting to set precedents about working in huggermugger

or in camera or whatever :P

hey guys — i have to jump to another meeting

stpeter, "in camera" - s'latin, innit?

Ashley - are you +1 on liason in general

yes!

k, more details to follow on the list, thanks for attending while you could

it sounds like there's some questions to be determined before we can really move forward in a meaningful way though

it sounds like we can respond "yes, but we have some followup questions"

yup yup

that seems reasonable

"yes in principle"

Yup, sounds good

so shall I work with the Board on a suitable reply and some follow-up questions?

+1

ok

+1

happy to have that discussion on the members@ list for transparency

+1

transparency is good

members@ works for this - will get more folks involved

I would appreciate if you could cc me any any correspondence

bear: yeah

feels awkward having to ask them what XSF sent to them, when being part of XSF and their working groups

lol

Peter Waher: if we discuss it on the members@ list, what we'll send will be in the open anyway :-)

i'm all for peter cc'ing the members list in all mailings TBH

(y)

bear: +1

super

should we meet again in a week to followup or will you need more time?

that should be fine — I think we can work things out on the members@ list

before then

but scheduling a follow-up seems fine

k, i'll email the list about a meeting next week

bear: great

peter - do we have more to discuss and do you have the needed information?

I have what I need and I don't personally have any other business, but we might want to ask if anyone else does :-)

:)

anything else to discuss/add ?

Not on my end besides looks like FOSDEM is all good

BTW, thanks to everyone for the input and questions about our topic of interest, that was all good feedback

I'm working out final details on a place to meet in Portland for the next Summit

so that should be nailed down very soon (this week)

the summit in portland should be very active

yes, lots of security topics of interest these days

:)

k, I think we can call today's meeting done then

Alrighty :)

thanks all for attending and thank you Peter for the work on this issue

Thanks all

See you next week

i'll do the minutes email after I get done yelling at some servers

great, I'll add the meeting to the calendar this time

thanks, bye

thanks Peter

time to head to the office, bbiab