From W3C Wiki

Derived from RRSAgent minutes


See also: IRC log


wilkie, evanpro, aaronpk, bret, +1.314.777.aaaa, rhiaro, tantek, AdamB, Arnaud, Lloyd_Fassett, jasnell, Sandro, oshepherd, markus, MarkC, bblfish, tiborKat


RESOLVED: Approve minutes of 30 September 2014


<trackbot> Date: 07 October 2014

<evanpro> OK, looks like we're rolling

<evanpro> Thanks hhalpin

<evanpro> Tsyesika: we just added you as an IE so you're welcome to join the call.

<evanpro> I've got T-7 minutes, is that correct?

<wilkie> yeah

<hhalpin> yep

<bret> o/

<hhalpin> Hey guys, I'm at a conference right now and its running late

<hhalpin> so I might be late to call. Evanpro, you are chairing, right?

<evanpro> hhalpin: correct

anyone on the sip line yet?

<hhalpin> My feeling is the best thing for this call would be to get commits to review from AS 2.0 from everyone

<hhalpin> and really actually do thorough reviews by next week - we still haven't had an IndieWeb review, for example.

<elf-pavlik> today I didn't manage to get VoIP working, so only IRC for me :(

<evanpro> My tab management issues around these calls is pretty severe

sorry about that, was just checking

<bret> hey tantek

<Loqi> tantek: bret left you a message on 10/1 at 3:03pm: I think it would be useful to have the JSON parsed out on each spec page, so I'll go start making modifications

<rhiaro> Zakim ??P7 is me

<tommorris> apologies for absence: work got busy.

<bret> PS, for anyone using the, the mute function does not work for really loud noises... be sure to mute with Zakim if you really don't want noise to come through

<wilkie> hello all

<jasnell> attempting to dial in...

<evanpro> That's fast

<Shane> Hi all, I'm afraid I am in public and forgot my headphones so won't join the call but I will stay in IRC :)

<Arnaud> who's ??P2 ?

<Arnaud> simple commands "topic: xxx" for agenda items, "xxx: blah blah" for when someone talks

<evanpro> scribe: cwebber

<wilkie> cwebber: there are some links to the commands you might find useful:

<wilkie> in particular

Approval of Minutes from 30 September 2014

evan says, let's get started, first order of business is to approve minutes from 30th of september

evan asks, resolving resolutions, are those handled to your satisfaction Arnaud ?


Arnaud: yes, but link is not correct, we should approve wiki page linked above (

which is the cleaned up version

evanpro says that looks good, we are using that version... do we have any opposition to approving the wiki page as minutes from previous meeting?

<sandro> cwebber, best format is like this: evan: whatever evan says

evanpro: RESOLVED: with unanimous silence, we've approved it

<sandro> (or evanpro or whatever)

(thx sandro)

<bret> yes

IndieWebCamp Cambridge


<wilkie> yes

<oshepherd> Wrong Cambridge, damnit! :p

<evanpro> :)

tantek: this is not an official w3c event, but members of the working group may be interested in an event at Cambridge, sandro and tantek are both co-organizers of this

<Loqi> aww

if you are interested in coming and working on your own personal website or etc come and join

cambridge MA, not Cambridge UK

<Shane> tantek: I will probably join remotely if I have time :)

cwebber: thx oshepherd, *, for helping ;)

evanpro: It's a good idea for us to stay in touch and keep these kinds of events on the agenda

<scribe> new agenda item: two main areas of discussion today. First is to have discussion based on our review of activitystreams 2.0 based on the last week. James has asked for feedback via github or the mailing list

we have open questions that have come up in discussion here

the second item to address is the social api issues

would like to discuss the social api ahead of TPAC, would like to ramp that up as we move forward with social context

also, want to discuss relationship with indieweb efforts

if we're going to lose tantek maybe we should move that particular item up to the front, is that okay tantek?

tantek: yep, can discuss for a few minutes

evanpro: great, if we talk the tough discussion early on, hopefully can keep it short, not take up too much of your time

Indieweb and JSON

<evanpro> elf-pavlik: we're discussing Indieweb and JSON

I think the person putting this online was elf-pavlik is participating only on irc

<evanpro> Can you characterize the issue quickly here in IRC?

<evanpro> elf-pavlik_: are you on?

evanpro: (repeating question) group charter of json serialization as a requirement

<elf-pavlik_> now yes, let me see onlin log

for the social data syntax

we have some ?? from indie social web community which may not support json

we also may need to discuss jsonl-ld, while most members seem to support, some indiewebcamp people may not support

clarification, we have *submissions* from indie sodcial web community

elf-pavlik brought this up, not sure we have actionable outcome from thi


cwebber: who's speaking?

<aaronpk> cwebber: that's tantek

tantek: what are we trying to explore?

cwebber: thanks, I'm still short on catching the voices

tantek: try to give the summary statement... in the indieweb community there's been really strong focus on doing absolute minimum you have to get things to work

there are so many implementers who are in their spare time doing here and there in their websites

following that path means minimum followed format

you just put up a web page with html, and from there

what's the minimum you can add to that to interoperate beyond that?

for blogposts, posts, etc

<jasnell> losing you tantek

cwebber: sounds like a plane is taking off

tantek: from there, what's the minimum to send back and forth between sites, and so pingback with even less work (without xml-rpc)

indieweb has gotten most of this working without any json on the wire

why do any more work if I don't have to?

very motivate to do the least they have to

it's not their job, they aren't some big enterprise

that said, the formats they're using have a canonical microformats reprsentation

so they can send through a canonical microformats representation

that's strictly a plumbing detail. They can also process with rdf if they want to

it's not opposition per say, they just haven't needed it

<jasnell> It's important to note that saying "The WG must at least produce a JSON-based serialization" does not equal "Everyone has to use the JSON-based serialization". Case in point: the AS2 Vocabulary can be easily represented using HTML5+(Microdata|RDFa|Microformat)

<elf-pavlik_> i have some concern about using application/x-www-form-urlencoded with complex nested JSON objects like AS2.0

evanpro: I have a bit of a disadvantage having person who brought this up not on call, only on irc

<evanpro> elf-pavlik_: will you be able to participate by voice next week?

<evanpro> Can we defer this issue?

<elf-pavlik_> evanpro, yes i will make sure

<Shane> I strongly agree, there are many (probably the majority) examples of social integrations between sites that use no javascript at all. Send using a form action, render on page using the serverside code, no JS or JSON required

evanpro: I'm not sure this issue is rooted in any of the work we're doing right now

<oshepherd> Shane: Nobody mentioned JavaScript

I'd prefer to deal with that when that comes arong


<Shane> oshepherd: Fair point

tantek: I agree

evanpro: so I'd prefer to do it when we have it grounded

<elf-pavlik_> +1 oshepherd

Activity Streams 2.0

let's move on to activitystreams 2.0 if no more discussion on this

activitystreams 2.0 is moving forward

james has been working on a new draft based on a structure of vocabulary and underlying syntax

as part of our work to evaluate that draft this week and next to make some decisions for that

<jasnell> first off... apologize for any background noise here if/when I speak

the point of having AS2.0 on the agenda is to have a chance to voice any questions and comments on the working draft

to see if there's anything we need to do to inform that discussion

if people can add folks to the queue if anything comes up


  • evanpro mutes out for a sec*

<evanpro> ack

jasnell: one of the main questions is whether we go forward and publish the current working draft?

as previous calls, the current draft is not expected to be perfect

it's just a statement of where we're going, what we're working on

two qs: is it ready to go? is it ready to go out?

second, are there any additional issues to resolve?

those are both separate questions, any feedback should be expressed in those terms

I will point out that there are some issues added to the github repo

<elf-pavlik_> +q can we add some issues around as:Link and as:Verb to draft inline?

in my views we have some technical questions added there, but no blockers

evanpro: great... oshepherd ?

oshepherd: at what point did we lose object type in favor of ?? type?

when did we switch from object type to json-ld's built-in (action??) type?

<oshepherd> @type


<wilkie> I'm less pedantic about key names and such as some of you :) but and I've reviewed it and it looks quite sound at the moment. my only concern/confusion was clarified earlier today. I might start implementing it. hmm.

jasnell: it's still there, basing on json-ld spect

previous versions not directly compatible with @type

so in order to achieve closer compatibility, aliased object type to alias of @type

one side effect is it's more backwards compatible with 1.0

but does use the value type concept I had previously

<oshepherd> Understood

one reference with one note sent to mailing list, I covered that previously

<bblfish> the mailing list? Which one?

evanpro: no problem, I have a question james
... elf-pavlik_ brought it up and put it on the agenda, we do seem to be strattling two compatibility issues, AS 1.0 and json-ld

can we do both, where do we err?

<elf-pavlik_> jasnell, can we add some 'bigger' issues inline to the draft? like around as:Link and as:Verb? also suggested in

jasnell: we can achieve 1.0 compatibility in that any 1.0 doc should be a valid 2.0 doc, but reverse is not true

we can decide that json-ld is the basis, and err on side of json-ld

if we have to break 1.0 compat to work better, json-ld it absolutely is

<elf-pavlik_> cool! :)

evanpro: if I'm not incorrect, there is recommendation to defer by media type

jasnell: that's correct, convention is that you walk 1.0 rules to apply... there's some questions about whether it's JSON LD plus (??)

cwebber: sorry, I lost a lot of that one

<oshepherd> Should probably add an editorial note discouraging serving as application/json

maybe a repeat if we want it better on record?

apologies, that last one I really garbled

jasnell: I do know it takes a couple of weeks to get the proposal draft going, so we need a decision quickly in time for TPAC

evanpro: yeah, last week we decided we would review for next 2 weeks on no/no-go decision on 14th

<jasnell> sorry chris: what I said was that the differentiation would occur on media type

we seem to be strongly leanign towards go, but we should make that decision by the 14th

<jasnell> application/stream+json for AS 1.0 rules

cwebber: thx

<jasnell> application/activity+json for AS 2.0 rules

evanpro: action item: everyone should review, be ready to make a thumbs up / thumbs down decision on putting forward a first public working draft

cwebber: who's speaking?


<evanpro> cwebber: it's Arnaud

<jasnell> Arnaud is speaking

Arnaud: yes, if you have any concerns, we should reviw current draft... if not, you need to bring up any issues that should be showstopper for you

which doesn't mean it's a perfect draft, but

evanpro: I'm sorry about that... people should be ready for overall structure/strategy of AS 2.0, if not every property name / structure

Arnaud: that's right


evanpro: great

Social API

unless there's anything more to discuss on AS 2.0, let's move on to the social API

our schedule, which we continue to slip behind on

I'd like to add action item for myself to update schedule

we are in a period where we should be starting to formulate our social api strategy

what we're going to do as our client->server api

we have at least one submission for open social activitystreams api

a RESTful CRUD api

we need to look at other APIs so we know what to look for

evaluating as we go in

tantek: I have to sign off for now, but there's a wiki page with other API candidates from charter / indieweb community

linked from social wg activity


talk soon, bye

evanpro: great, thx a lot tantek

as for social api candidate someone posted, we talked a little bit about opensocial / activitystreams api

I think there are high level qs: how broad of an api, how much do we want to spec out, do we want to keep it lean/small?

there are some social API patterns listed, not sure they match, some of them are inappropriate / proprietary

none of facebook/twitter/google have made submissions

but may be good patterns to understand: what is a social api?

<hhalpin> I suggest we try to loop back in OpenSocial, as they seemed to have some sort of plan but haven't been attending phone calls.

as an outcome of this meeting, maybe put on our schedule for tpac: an action to come out with our social API strategy from our meeting

bblfish: I'm looking at this social api, what I don't see is LDP up, that's very generic

<jasnell> We need a definition: What does a "Social API" do?

is that at the wrong level, or at the right level?

evanpro: you could add it, give a link for those of us who don't know it, that would be helpful

bblfish: ok

<wilkie> I think extensibility is a priority. argh. wish I could afford to get to this TPAC meeting.

evanpro: jasnell would you mind bringing that to voice?

jasnell: yes, post it on the ml?

evanpro: no, on irc, what does a social api do

jasnell: yes, as for strategy, what does a social api *do*? we can make all sorts of examples, but how do we evaluate what we're trying to do

I don't know what we're trying to achieve, how do we know what we try to do?

evanpro: yes, we're trying to make a list of patterns that meet the same use cases as these apis

<tiborKat> zakim : +tiborKat


google, twitter, open social api... may make some sense to go finer than that

to say these use cases we're trying to address

<bret> Zakim: who is making noise?

I also tried to throw on a wiki page but

there are also some strategy issues: we have lean api from opensocial, but may be more we want to address

I want to add that we want to be ready by TPAC time to talk about strategy for APIs

<evanpro> +1

would like to take a little poll to see if that's reasonable

<tiborKat> yes I'm on IPCall

that's for end of october to say strategy of: simple the opensocial activitystreams api, or opensocial activitystreams api + endpoint, will we be ready for that?

<jasnell> +1

<wilkie> +1

give me +1, -1, 0 on IRC

<oshepherd> +0.7

<bret> +1

<AdamB> 0

<markus> +0

<rhiaro> +0

<MarkCrawford> +1

<elf-pavlik_> 0



<Shane> +0




evanpro: I'll take that as net 1 but pretty weak net 1

<jasnell> I think we need to at least schedule time to brainstorm/discuss API case proposals at TPAC

<wilkie> strawpool to figure out who all the optimists are :)

<jasnell> then see what people think of those

this is something we should discuss pretty extensively at tpac

<Shane> wilkie++

<AdamB> +1 on discussing +0 on being ready for a strategy

<Arnaud> +1 to discussing this at TPAC

<Loqi> wilkie has 1 karma

I like that structure jasnell, to discuss proposals at tpac

come up with strategies by end of tpac

<Shane> +1 to discuss at TPAC, though I won't be there

okay! I'd like to ask group to review wiki pages that have come up, add to them, discuss on list if you have questions

<jasnell> hopefully we'll have an initial set of use cases proposed from the Social IG by then also

<elf-pavlik_> +1 to review proposed APIs

once we get to our social data impact, we'll need to be moving fairly quickly with our social apis

I believe that's the end of our scheduled agenda

anything else to discuss today?

add yourself to queue if anything to bring up

<Arnaud> +q

if not, we're ready to abjourn

Arnaud: quickly, yet another reminder, TPAC registration deadline is tomorrow before prices go up

evanpro: okay great, thanks everyone for coming

see you all next week

<elf-pavlik_> thanks evanpro & cwebber !

<evanpro> cwebber: OK, hold on one last part


<evanpro> To generate the minutes there's another invocation you have to do

evanpro: ok

cwebber: ok

<oshepherd> Also, I suggest using PanDoc to do a quick/rough HTML-to-Wikitext conversion

and yw elf-pavlik_, hopefully I didn't do too bad ;)

<aaronpk> cwebber++ for excellent scribing

<Loqi> cwebber has 1 karma

<elf-pavlik_> cwebber++

<Loqi> cwebber has 2 karma

<Arnaud> evanpro, you just need to end the meeting

<Arnaud> trackbot will do the rest

<trackbot> Sorry, Arnaud, I don't understand 'trackbot will do the rest'. Please refer to <> for help.

<evanpro> trackbot, end meeting

cwebber: evanpro: what's the invocation I have to do

Summary of Action Items

[End of minutes]