Social Web Working Group Teleconference -- 09 Sep 2014

<trackbot> Date: 09 September 2014

<evanpro> Thanks, I didn't do that yet

<evanpro> no, akuckartz wont be here today so we need a new scribe

<tiborKatelbach> Zakim: +tiborKatelbach

<Loqi> tiborKatelbach: tantek left you a message on 8/22 at 1:56pm: yes, telecon dialing in etc. information is at https://www.w3.org/wiki/Socialwg#Telecons and be sure to be on IRC as well for minutes, additional converstaion, queuing.

<evanpro> Arnaud: you're next alphabetically, would you mind scribing?

<deiu> Regrets on my part (busy with something)

trackbot, start meeting

<trackbot> Meeting: Social Web Working Group Teleconference

<trackbot> Date: 09 September 2014

<evanpro> Uh, again?

<evanpro> I thought it was already started.

it was, evanpro -- my mistake

<evanpro> NP

<evanpro> One thing we need is a scribe

it's just it didn't identify the telecon properly

<evanpro> Actually, I'll just start

<jasnell_> joining in a second

<evanpro> I'm one of those IP callers

<aaronpk> Zakim: aaaa is aaronpk

<jtauber> i may be IPcaller.a

<evanpro> Tada

<evanpro> elf-pavlik: is that you?

<tiborKatelbach> zakim +tiborKatelbach

<tiborKatelbach> zakim aabb is me

<Tsyesika> i'm not on the call btw :s not sure why it thinks i am :P

<dromasca> ??P15 is me

<harry> Sandro :)

<evanpro> Perfect

<evanpro> sandro: can you scribe for us?

<harry> Tysesika did it last time

<harry> Evan is chairing

<harry> jasnell has to speak a lot this meeting I imagine

<Tsyesika> i did what

<Loqi> evanpro meant to say: sandro: can you scribe for us?

<jasnell> I will volunteer to scribe next week

<tommorris> Zakim: IPcaller is me

<evanpro> Arnaud chaired twice in a row

<scribe> scribe: sandro

<harry> ACTION: harry to create a scribe list [recorded in http://www.w3.org/2014/09/09-social-minutes.html#action01]

<trackbot> Created ACTION-1 - Create a scribe list [on Harry Halpin - due 2014-09-16].

<evanpro> https://www.w3.org/wiki/Socialwg/Scribes

evanpro: working on scribes list, but it's not quite done yet

<evanpro> https://www.w3.org/wiki/Socialwg/2014-09-02-minutes

evanpro: minutes from last week

PROPOSED: approve minutes from last week https://www.w3.org/wiki/Socialwg/2014-09-02-minutes

RESOLUTION: approve minutes from last week https://www.w3.org/wiki/Socialwg/2014-09-02-minutes

evanpro: this is the 3rd official meeting. we're behind schedule, for various reasons
... I've revised the schedule. It's ambitious, but I think we can meet it.
... I'd like to run throught he schedule

<evanpro> https://www.w3.org/wiki/Socialwg#sched

evanpro: For September, our big item is picking an initial social data syntax
... with the goal of having a draft for TPAC
... with the intention of by next week we'll have an agreed set of candidates and criteria
... and two weeks from now we'll have confirmed that direction
... and then can begin making a draft.

evan: I'm an AS advocate, but I want us to consider other options

<evanpro> Zakim: q?

<Zakim> sandro, you wanted to ask about draft publication schedule

<harry> either publication at TPAC or before TPAC is fine with me.

<harry> It does tend to help focus discussion.

sandro: goal is to have consensus to publish BEFORE of AFTER tpac?

evanpro: probably after

<jasnell> My proposal is currently documented in the AS2 draft... specifically... JSON + AS2 Vocabulary + (JSON-LD alignment + recommended use of Vcard, Foaf, Org Ont, DC, and Prov ontologies). Other ontologies such as schema.org *could* be used, but would not be specifically recommended

<jasnell> AS2 working draft is located here currently: http://jasnell.github.io/w3c-socialwg-activitystreams/activitystreams2.html

<evanpro> Thanks

<jasnell> can snapshot that at any time to product FPWD

elf-pavlik: Comparing AS and Schema.Org

evanpro: Topic at the moment is the schedule itself.
... AS / schema discussion is later on the agenda
... Is the existing list of candidates and requirements reasonable?

<Loqi> sandro meant to say: evanpro: Is the existing list of candidates and requirements reasonable?

evan: And can we make the decision within two weeks

evanpro: Hearing no comments, I'm going to take that as agreement

shepazu: As a way of getting feedback, you might ask for a specific proposal

<elf-pavlik> this one https://www.w3.org/wiki/Socialwg#sched ?

<wilkie> elf-pavlik: yes

<evanpro> elf-pavlik: yes

PROPOSED: The schedule for the Social Data Syntax is as just discussed, closing out candidates & requirements next week, trying to for dececision in two weeks, and draft published after F2F

<shepazu> +1

<harry> +1

<wilkie> +1

<evanpro> +1

<tiborKatelbach> +1

<dromasca> +1

<elf-pavlik> +1

<jtauber> +1

<oshepherd_> +1

<DaveSkiba> +1

<jasnell> +1

<Arnaud> +1

RESOLUTION: The schedule for the Social Data Syntax is as just discussed, closing out candidates & requirements next week, trying to for dececision in two weeks, and draft published after F2F

shepazu: Because this group has decided to work asynchronously, was there a call for consensus on this on the mailing list as well?

evanpro: No, so there should be

<elf-pavlik> do we have this workflow documented on our wiki?

shepazu: How about: Report this proposal to the mailing list, showing consensus on mailing list, and please respond to let us know what you think (giving a deadline). This way you've checked all the boxes in terms of being async.

<harry> How about RESOLVED = what shepazu just said

<harry> +1

Arnaud: This sounds like a heavy process. The minutes are published -- this should stick until/unless someone raises a concern on the mailing list.

<oshepherd_> Perhaps a summary of resolutions in the minutes E-Mail?

shepazu: When groups I've been in have said they want to work asynchronously, they've accepted the idea that people wont be on calls. If you want to say people have to read the minutes to participate, that's fine, too.

<tiborKatelbach> The group could vote for Arnaud or dougs view

<elf-pavlik> +1 doug

Arnaud: Since we've decided to put the minutes in the wiki, we can highlight the resolutions at the top
... then they stick a week later, when minutes are approved

<tiborKatelbach> +1 doug

shepazu: There's a perl tool that pulls out the resolutions and writes them to a wiki page for the SVG wg

<wilkie> whhaaa

<wilkie> darn it all! nobody told me that last week haha

<harry> we have no official tools

shepazu: You can use whatever tools you want

<harry> Resolutions can be objected within a week.

evanpro: Sounds like -- bring resolutions to top of minutes, but don't send special email. And people have up to the week later.

<harry> That sounds rather reaosnable

<elf-pavlik> 0

<wilkie> +1

<tiborKatelbach> +1

<harry> +1

<evanpro> +1

<Arnaud> +1

<jtauber> +1

+0 wishing for better tools, but yeah

<shepazu> +1

<jasnell> +1

RESOLUTION: resolutions will be highlighted in minutes, and people not at meeting can bring up their concerns within the following week

evanpro: social client api -- collecting candidates and requirements
... decision by mid to late october

<harry> sounds reasonable

PROPOSED: Social API schedule: gather candidates and requirements by mid October

<elf-pavlik> +1

<oshepherd> +1

<tiborKatelbach> +1

<wilkie> +1

<harry> I'd like to see what other AS 2.0 work besides OpenSocial is in this space

<harry> _

<harry> +1

<jasnell> candidates and requirement *proposals* right?

<harry> yes

<dromasca> +1

PROPOSED: Social API schedule: gather candidates and requirements by mid October, with confirmed direction shortly thereafter, but no draft approved at TPAC

<Arnaud> +1

<DaveSkiba> +1

<jtauber> +1

<evanpro> +1

<jasnell> +1 then...

<elf-pavlik> +1

RESOLUTION: Social API schedule: gather candidates and requirements by mid October, with confirmed direction shortly thereafter, but no draft approved at TPAC

<harry> I think re federation that's pretty hard - we can modify schedule later.

<harry> Focus on API and syntax first.

<elf-pavlik> +1 harry

evanpro: Server-to-Server / Federation protocol, I'd like to put that off a bit, until we have some bandwidth to look at them, then look at schedule.

<tiborKatelbach> :)

<harry> Anyways, just don't worry - we can modify agenda in charter

<harry> no big deal

<harry> We can move that back

<evanpro> http://www.w3.org/2013/socialweb/social-wg-charter.html

<harry> to later in 2015

<oshepherd> A point I've raised in the past: I think that any comprehensive social API and federation protocol are fundamentally intertwined concepts

evanpro: I think federation in the charter is Q1 or Q2 2015

oshepherd, do you want to voice that?

<oshepherd> Sorry, my mic is still out of commission

evanpro reads oshepherd's line

<jasnell> a properly designed API will also handle the federation requirement

<jasnell> I don't view those as separate items

evanpro: For example ostatus and pump.io many of the rest endpoints are used for both, so it probably makes sense to consider social api

<elf-pavlik> sandro: federation protocols first and client api just special case (elf-pavlik tries help scribing)

evanpro: There are many social APIs in use across the web, and we can provide value by standardizing that, without addressing federation. I wouldn't want to delay that in the name of thinking about federation.

<shepazu> I'm curious how the different options for Social Data Syntax and Social API are going to be evaluated… are there use cases and requirements for these?

evanpro: we do have a submission (embedded experiences), and ?? from open social, which could work very well for social API

<evanpro> http://www.w3.org/Submission/2014/SUBM-osapi-20140314/

<elf-pavlik> shepazu, i poposed use case https://www.w3.org/wiki/Socialwg/Check-In_Use_Case

<evanpro> https://www.w3.org/wiki/Socialwg/Social_syntax_requirements

evanpro: Answering shepazu -- I think we're talking about collecting them for discussion. We do have a social data syntax requirements page on the wiki, but they are very loose and informal.
... We could get pretty far down the hole in use case analysis

elf-pavlik: it's hard to evaluate without use cases.

<harry> I find use-cases tend to blackhole in this space, look for running code people use instead.

elf-pavlik: this came up in considering Hydra

<harry> For example, the Social XG made a 50 use-case document that was basically infinite and never finished.

evanpro: 95% of use case analysis is in the activity streams already

<harry> http://www.w3.org/2005/Incubator/socialweb/wiki/FinalUseCases

<harry> use-cases in this space are very large

<harry> I'd recommend looking for *running code* with *real users*

<harry> And see what they are *actually doing* - and include both open-source and proprietary systems in this codebase analysis.

evanpro: current wiki list is very loose; if we did use case justification, it would go on and on.

<harry> For example, IBM has Connections, Evan has pump.io, SugarCRM, SAP Jam, IndieWeb sites

<harry> I'd like to see a list of products and software rather than use-cases, and then look maybe at the problems that real running software is trying to solve.

<elf-pavlik> harry, also Crosscloud :) http://cimba.co/ | http://www.w3.org/2013/socialweb/social-wg-charter.html

<harry> Since 90% of the software in the "decentralized" social space is essentially vapourware

tommorris: Looking at use cases for federation protocol, and wondering how it relates to charter requirements. Are we talking about an individual social network (eg facebook or twitter), or .... what are the servers we're federating?

<harry> Federation = HTTP REST-based protocol for federating data, including obvious things like blog posts but possibly also personal data.

evanpro: I think the intention is to have multoiple servers under the control of different entities, including consumer oriented social networks and enterprises, able to share data between servers, hetergeneous, like SMTP

<jasnell> Part of the challenge here is that we are not working off a common definition of what a "Social API" means... or what "Social Federation" means... we need proposed definitions

<harry> I think definitions will rathole just as bad as use-cases. Instead, look at working software that claims they have federation or want it.

evanpro: It would be less interesting and not worth our energy to look at how folks should do these things internally.

<harry> Ditto API.

dromasca: (from Avaya) This definition of federation you just described, this is much beyond social API. it brings up data ownership. not sure that's in current charter.

<harry> Data ownership is probably a separate topic but we should help enable that.

<oshepherd> dromasca: Data ownership will probably be similar to SMTP

elf-pavlik: I think it would help to discuss things more on the mailing list

<oshepherd> dromasca: I.E. if you send it to somebody, you probably can't get it back

<elf-pavlik> discuss things *before* teleconf

evanpro: We're putting Federation until after TPAC; concentrate on client api and social data standard now.
... can we table this Federation talk for now?

<evanpro> +1

<elf-pavlik> +1

<harry> +1

<wilkie> +1

<jtauber> +1

<dromasca> +1

<shepazu> +1

evanpro: (except where it might require things of the Social API)

<jasnell> +1

<DaveSkiba> +1

<tiborKatelbach> +1

Social Data Syntax

evanpro: We have some candidates on the wiki, as well as some loose requirements
... Hydra, Schema.org, AS2 are all strong candidates
... We also included some not-submitted structures, like OGP and Tweets
... Facebook is not interested in submitting OGP
... or licencing it for our use
... There are some others on here

<harry> Twitter is discussing us with Friday

<harry> will send email after that!

<harry> Twitter is new to open standards space

evanpro: Unlikely they'll submit that, but it's possible

<jasnell> The syntax ought to be: JSON + JSON-JD + minimal set of vocabularies including AS2 and the others documented in the AS2 draft. It should be possible to use schema.org, but it should not be a recommendation of the WG to use schema.org unless schema.org governance is opened up.

elf-pavlik: Hydra may start using the schema.org vocabulary

<jasnell> it's great that schema.org updated terms of service, but governance is still a challenge

<tiborKatelbach> +1

<Arnaud> schema.org is not a syntax though

<harry> The previous good argument was that schema.org did not conform with RFF patent policy

jasnell: There's the syntax and the vocabs. For the syntax, if we focus on JSON and JSON-LD, that covers that. For Schema.Org, it's fantastic they changed the terms to be compatible with W3C, but governances is still an issue.

<harry> I think the main argument is how to harmonize so we can use schema.org nicely with something like AS 2.0.

jasnell: While that's still the case, I don't think W3C should be recommending people to go use it.
... It's possible for individual developers to use from multiple vocabs

<oshepherd> There should also be concern over Schema.org's quality control

jasnell: so it's possible to use schema. org but util we work something out, we shouldnt recommend

<evanpro> sandro: although schema.org will not turn over the vocabulary entirely it could be dual-homed

<dromasca> +q

sandro: maybe we can align terms between the two

<Arnaud> +q

evanpro: if schema.org meets our basic requirements, I'd like to at least consider it on technical grounds. I'm not sure if it does. I'd like to analyse it from those grounds, before we put a lot of effort into dual-homing

<harry> For schema.org we'd also need Google in WG to make any parts of it move forward on Rec track BTW

dromasca: It's probably too early, since you're right about technical issue need to be addressed first.

<jasnell> we need to separate the syntax and vocabulary discussions a bit. The syntax is the JSON, JSON-LD and basic encoding guidelines (much of what AS2 defines)... the vocabulary is the data model, the set of property names, the semantics, etc

<jasnell> we can decide on the syntax separately from ironing out all the vocab issues

sandro: No promises on governance, but I think it's probably possible, based on talks with DanBri, etc.

Arnaud: So we could copy schema.org and put it under our own control? And if everyone plays nicely, things keep working.
... Also, it seems like this wiki page is a good start. We need to separate syntax from vocabularies.
... Can we separate candidates, apples from oranges

<oshepherd> +1 evanpro

evanpro: From my point of view -- we're not talking about all of schema.org, just the Action section. Separating syntax from vocabs could really confuse users, so I'm wary about separating them.
... We've hit the top of the hour. Can we continue another 15 minutes?

<elf-pavlik> +1

<wilkie> +1

<tiborKatelbach> +1

<harry> +1 (but without me)

<evanpro> +1

<jtauber> +1

<AdamB> +1

<oshepherd> +1

<Arnaud> +1

<oshepherd> I think there is a need for a *simple*, *concise* vocabulary with a tight relation to the syntax

evanpro: any more discussion

<oshepherd> Schema.org vs JSON-LD are orthogonal matters

<tiborKatelbach> +1

<evanpro> +1

<oshepherd> +1

strawpoll: json-ld plays a part here for us

<wilkie> +1

<elf-pavlik> +1

<jasnell> +0.5 ... we need to be clear... AS2 is JSON that is compatible with JSON-LD

<elf-pavlik> jasnell, can you write to mailing list about difference?

(I see that as json-ld playing a part)

<tommorris> +0 - if RDFish Linked Data matches the use cases, JSON-LD seems a natural technology to use.

<jasnell> elf: yes

<oshepherd> (I'll note that jasnell's point is the entire purpose of JSON-LD)

<tiborKatelbach> schema.org ontolgies can also be written in json-ld

jasnell: JSON-LD implies both the JSON syntax and a particular processing model, and the @id kind of stuff. For backward compatibility, AS2 did not go there.
... AS2 does not say that it is JSON-LD, but here is a set of guidelines for how to process it as JSON-LD.
... Important distinction

<elf-pavlik> i understand backward compatibility to AS1

evanpro: My hope is that we understand that well before the F2F

<oshepherd> I'd like to see the vocab and processing model properly separated from the syntax portions of the AS2 spec

evanpro: jasnell would you take an action to describe that structure?

<scribe> ACTION: jasnell to describe how AS2 diverges from JSON-LD and manages the compatibility [recorded in http://www.w3.org/2014/09/09-social-minutes.html#action02]

<trackbot> Error finding 'jasnell'. You can review and register nicknames at <http://www.w3.org/Social/track/users>.

<oshepherd> (A lot of the issues I've raised with AS2 are because the mapping is not at all clear)

<jasnell> jsnell

<scribe> ACTION: jsnell to describe how AS2 diverges from JSON-LD and manages the compatibility [recorded in http://www.w3.org/2014/09/09-social-minutes.html#action03]

<trackbot> Created ACTION-2 - Describe how as2 diverges from json-ld and manages the compatibility [on James Snell - due 2014-09-16].

evanpro: Furthur discussion?
... continue this on the mailing list, please

<elf-pavlik> i can take action with help of others: pelf

sandro: Anyone willing to compare AS2 and schema.org, so we don't have to all do it for ourselves??

<tommorris> Wiki-page?

<jasnell> I've documented a few comparisons on my personal blog in the past, will provide links in the wiki

<scribe> ACTION: elf to work on comparison of AS2 vocab and schema.org actions vocab [recorded in http://www.w3.org/2014/09/09-social-minutes.html#action04]

<trackbot> Created ACTION-3 - Work on comparison of as2 vocab and schema.org actions vocab [on Pavlik elf - due 2014-09-16].

tiborKatelbach, I'll help: too!

<jasnell> and will post a few of my own comments to the mailing list as time allows

<Loqi> sandro meant to say: tiborKatelbach: I'll help: too!

elf-pavlik: On social syntax vs vocab, the way collections work, Hydra and LDP do that differently. Itemlist on schema.

+1 collections in LDP are interesting

<elf-pavlik> https://www.w3.org/community/hydra/wiki/Collection_Design

sandro: LDP containers *can* impact your vocabulary desing

<oshepherd> sandro: I'm not quite sure how collections in Hydra are any different from OWL

<elf-pavlik> +1 mailing list

<jasnell> +1 to documented proposals on the mailing list

sandro: Let's take that to the mailin glist.

<jasnell> who is beeping...?

Organizational Topics

evanpro: trying to get agenda out before the weekend --- I'll try

Arnaud: Yeah, I think I can do that, too.

<elf-pavlik> thanks :)

<evanpro> +1

<elf-pavlik> +1

evanpro: folks like this?

<wilkie> +1

evanpro: might miss current conversation

Arnaud: Friday before should be fine.

evanpro: Presenting social efforts during other events

elf-pavlik: I'm going to some events; I'll work on a slide deck. Anyone have slides, or want to help?

<elf-pavlik> yes

elf-pavlik: I'll send email

evanpro: We're out of time. Any other business?

ADJOURN

thanks Evan!

<oshepherd> :-)

<elf-pavlik> thanks Evan & sandro!

<evanpro> trackbot, end meeting

<elf-pavlik> i think P19 was me

trackbot, end meeting

- DRAFT -

Social Web Working Group Teleconference

09 Sep 2014

Attendees

Contents

Social Data Syntax

Organizational Topics

Summary of Action Items

Scribe.perl diagnostic output