Social Web Working Group Teleconference

16 Jun 2015

Attendees

Present: Sandro, Arnaud, aaronpk, Ann, +1.773.614.aaaa, rhiaro, +1.401.305.aabb, ben_thatmustbeme, cwebber2, +1.857.445.aacc, bblfish, +1.314.777.aadd, AdamB, hhalpin, KevinMarks
Regrets
Chair: Arnaud
Scribe: rhiaro, aaronpk

Arnaud: When we had the f2f we decided it would make sense to converge the proposals and start to draft a document
... Starting with a strawman, then fill in the blanks
... Look at the different proposals, have some alternatives and see which we prefer
... There are several people who volunteered to be editors
... Aaron, Amy and Jessica

<hhalpin> I think Tysekia also volunteered to help

Arnaud: So, update? Is there progress?

<Arnaud> ack: aaronpk

<hhalpin> In particular, any unresolvalbe differences?

Arnaud: This is an opportunity for everyone to get on the same page

aaronpk: We've been making slow progress
... What I'm trying to do is base it on actual implementation experience
... Been slow because I have a backlog of things to do
... However there's now an updated version of the brainstorming document

<aaronpk> https://github.com/w3c-social/Social-APIs-Brainstorming#micropub

aaronpk: Updated micropub section
... I implemented this for creating, editing and deleting posts

<AnnB> big thanks to rhiaro, aaronpk, tsyesika!

aaronpk: Plan is to use that as my new micropub endpoint on my own site, had a good experience following this spec
... There are two ways to do each post, one form encoded, one json
... If you take a look at that, you see examples of both versions
... Still more to do to demonstrate this in a useful format, because all it's doing right now is modifying my storage files, not generating output anyone else can see yet
... This syntax is based on the idea of edit-posts

<aaronpk> http://indiewebcamp.com/edit

aaronpk: The idea that an operation like an edit can be represented with its own URL and the chnages can be described at that URL using microformats
... More background on that wiki page ^

<aaronpk> i can scribe for amy

<aaronpk> scribenick: aaronpk

<rhiaro> https://github.com/w3c-social/activitypump/blob/master/implementation.md

rhiaro: started going through the activitypump spec and rewrote it in a checklist form ^
... understand how to implement activitypump and how it maps to micropub and the others, micropub is basically a subset of AP, as well as webmention is
... next steps are going through micropub and seeing what's missing that can be achieved with activitypump and see how they can fit toghether that way
... that might be the basis for this document we were talking about
... haven't gotten much feedback from jessica on that document yet, would like to hear from her soon

<melvster> are we agreed on implementing inboxes and outboxes? SoLiD doesnt do that yet, but we've discussed it ...

Arnaud: right now we have different documents describing the two approaches, but still need to get started on pulling them together into *the* spec

<AnnB> is someone going to do the comparable work with linked data perspective?

<cwebber2> +q

AnnB asked the question she typed in IRC

bblfish: haven't looked at this yet but had some questions and will send them on the list
... i think melvin might have been following closer

<AnnB> ... or Andreas

Arnaud: you will need to figure out who there will step up to do that

<rhiaro> scribenick: rhiaro

cwebber2: The convergence doc is affecting how both micropub and activitypump documents are developing
... From seeing forechannel and backchannel discussions between Jessica, Aaron and Amy

<AnnB> dog: bark, bark

cwebber2: I'd like to see strawman document, but it does seem to be affecting real things

<AnnB> :-)

<cwebber2> sorry I couldnt's mute my phone

Arnaud: How long is it going to take before we can try to have this document?

<cwebber2> sorry about the barks

Arnaud: Investigate further what's possible, what's missing etc
... Sounds like people are busy with prep work

<aaronpk> scribenick: aaronpk

<hhalpin> From W3C's perspective, we'd want a FPWD by TPAC at latest - we're quite behind schedule :)

rhiaro: we could start putting something together this week, would be a starting point

<hhalpin> Whenever the editors are comfortable, quite happy to do a Respec.js training and set them up in CVS etc.

<AnnB> hhalpin, are you going to dial in?

Arnaud: i understand the challenge of doing this but it was harry's suggestion that we try and mgiht just be a strawman
... but that would give us a sense of what needs to be specified

hhalpin: i agree with amy that we don't want to create yet a third or 4th document that is a fork of the practical work being done
... my proposal would be whenever folks are comfortable to converge at least two of the docs, we can do the first editor's draft
... i would jhust say whenever you guys are ready ping the WG and we can start the editor's document

(discussion about github vs cvs editing flow)

Arnaud: this is a new tool that makes publishing in TR space as easy as it can possibly be

<hhalpin> See here for links:

<hhalpin> https://github.com/w3c/echidna/wiki/How-to-use-Echidna-with-ReSpec-and-GitHub

Arnaud: which is a lot lot better than the old process

<hhalpin> Much *easier*!

<hhalpin> Here's the other link:

<hhalpin> https://github.com/w3c/echidna/wiki/How-to-use-Echidna

sandro: is aaron's new json document enough to implement this?

rhiaro: the brainstorming doc is a comparison of activitypump and micropub, both are implementaable. both do things that the other doesn't do, and use slightly different vocabularies

sandro: both of you said you're implementing before you write more down

<rhiaro> scribenick: rhiaro

aaronpk: the document there right now is part of the picture, the micropub syntax for creating things
... you can implement it, but it doesn't have the other half of it which is rendering html, what you do with it when you get the commands
... because that's where I am in my implementation process
... I've been writing the document, then writing the code, and going back and forth
... So it'll keep progressing that way for a while
... Part of the goal of the brainstorming document is pulling in the things from activitypump that I'm going to need to do that aren't in micropub yet
... that's where the convergence is happening
... I find it more useful to have code alongside it to justify things in the spec

AnnB: Sounds great. Amy proposed to write a strawman document, although harry said we don't need another document. It sounds like what you're working on now are technical explorations, and maybe it is valuable to have a comprehensive draft of what the ultimate document might be like

<hhalpin> Ann, to be precise I said there was no need to rush it, but I'm happy to set-up an editor's draft when Amy, Aaron, and Tysika are ready

Arnaud: I second this, I thought that was the plan
... Current documents are useful in documenting the status quo, but not necessarily what we need in the end
... So you could keep working on that and slowly morph it into the spec that we want to produce
... Might be easier to start a fresh document that is meant to be the final spec, and have the outline and fill in the blanks
... It's a matter of editorial method
... Happy for editors to figure out what works best

<AnnB> I agree .. it's up to the editors, which way works for you guys

bblfish: As I understand, it's going very deeply into how you can tie activitypump and micropub together. Before you go to deep it would be helpful to try to do the opposite with activitypump and linked data. Start with major alignments and see if the pieces fit together
... How would one do that

AnnB: We're waiting for someone from the linked data side to do that work

Arnaud: go ahead!
... Someone needs to step up, they are willing to accommodate this, but they can't invent the information. Someone from SoLiD needs to step up and provide this

bblfish: In order to do this one needs to work at the big pieces first, see how they align

<hhalpin> It's also fine to go linearly - i.e. see how activitypump and micropub goes together, and then align that result with SoLID

<hhalpin> whatever works for folks

bblfish: That you can do by writing up the different pieces that exist and trying to tie one to the other

<AnnB> great, bblfish ... do it!

<cwebber2> bblfish, that's what the brainstorming doc is doing already

bblfish: Then you can have a discussion

<cwebber2> but it needs help from solid

bblfish: Wanted to check that everybody agreed this is the right proces

AnnB: it seems like you or melvster or any of the other SoLiD people just need to start doing something

Arnaud: what we're trying to tell you is there is no disagreement with this approach, there's a lack of putting things in writing from the linked data side

bblfish - please do make a PR to the brainstorming doc to fill in the same things fro SoLiD!

<hhalpin> Well, I think Andrei is updating the SoLID docs

<aaronpk> rhiaro: that document is what we've been doing, comparing activitypump and micropub, we'd love to get SoLiD in there as well. just make a pull request, we'd be happy to see that

https://github.com/w3c-social/Social-APIs-Brainstorming

<aaronpk> bblfish: which document?

<ben_thatmustbeme> bblfish: its in the agenda too i believe

<bblfish> thanks

<scribe> scribenick: rhiaro

sandro: I'd like to encourage LDP people to try the implementation first approach

<melvster> sandro++

<Loqi> sandro has 14 karma

bblfish: in LDP we had the implementations a year before the WG

Arnaud: you can do both approaches at the same time
... aaron is trying to document and implement and going back and forth, you could do something similar
... What sandro is saying is we don't want just vapourware, we want to ground the work in real implementations

sandro: the SoLiD code works, there are a bunch of servers and a bunch of clients, we could try to get them to speak some elements fo activitypump and micropub and see what changes that requires (just thinking aloud)

Arnaud: hopefully people have a better understanding of where we stand

<ben_thatmustbeme> yes

Arnaud: Is ben_thatmustbeme on the call?

<ben_thatmustbeme> very tired

AS2 language support

ben_thatmustbeme: Jessica would also have a lot to say and isn't here
... Basically I've been going through AS2 and fixing microformats examples to match real world implementation
... Highlighting parts where I question whether they're needed or if microformats needs to add features
... One that did come up was multi-language support

<cwebber2> +q

ben_thatmustbeme: in AS2 you can add multiple translated versions of the same post

<rhiaro> (contentMap)

ben_thatmustbeme: I want to float the idea of saying that it doesn't match any user story, no discusison anywhere of multiple languages

<hhalpin> +1 multiple languages

ben_thatmustbeme: And discuss possibly removing that

<melvster> FYI: (not just *post*, any text)

ben_thatmustbeme: But without james and without jessica I don't know how far we can get

cwebber2: I have a suggestion that might make things easier

<hhalpin> Note that W3C does have staff that can help with this

<hhalpin> if we need some advise, i.e. Richard Ishida

cwebber2: The handling things in multiple langues is already handles in json-ld
... It might simplify things if we don't mention it in the AS2 spec, but don't exclude it either
... It's still possible via json-ld
... But we don't actually specify it in the spec, so we can kind of explore it more in the future
... But it doesn't necessarily set up a barrier when we don't know how hard it's going to be
... I do agree that we don't need to conclude anything while jessica isn't here

<KevinMarks> language is also included in microformats implicitly as html has the lang attributr

<rhiaro> I also assume elf would have something to say

harry: w3c actually has i18n staff who can give us advice

<KevinMarks> though currently not part of the mf2 parsing spec

<cwebber2> sorry, spotty connection as I said

harry: Sounds like we can lean on json-ld, but if we need it, the w3c has people whose full time job thi sis

<cwebber2> I'll ust have to participate via text for now

harry: I agree that multiple language support is a good thing for specs in general

<hhalpin> http://www.w3.org/International/

Arnaud: from the i18n working group, we should have this

<hhalpin> If we have any questions, W3C has specialist staff can help

sandro: json-ld inherits from rdf a halfway decent notion of multi language stuff
... but it doesn't have really clarified is 'is a given value available in multiple languages'
... eg. does a particular city have a name in english and a name in french and a name in russian.. does that count as multiple values, or one value with multiple tags
... json-ld isn't going to solve that for us
... we need to solve or worry about it later
... the i18n activity doesn't know anything about this either
... they're happy to work through use cases like ltr vs rtl script and we'll have to get them to review that
... but they don't know anything specific about these technologies that would be helpful

ben_thatmustbeme: Just want to be clear that I understand setting a language and saying this text is english, this is an entire post drafted in english then draft the same post in another language
... and have that be the same object

sandro: then it gets weird because your objects model gets confused about how many posts that is

<KevinMarks> rel="alternate" hreflang="fr" etc works for whole posts

<AnnB> aaargh

<KevinMarks> it's mixed language within a post that is trickier

<cwebber2> brb

Arnaud: We should have the possibility of making posts in different languages, but not necessarily handle this case of multiple versions of the post in multiple languages

ben_thatmustbeme: yes
... On facebook/G+ this isn't a feature. I have friends who will include two languages in one post

<cwebber2> back

Arnaud: how is ti marked up?

ben_thatmustbeme: it's not

sandro: some languages you can't tell what it is by looking at it, but a screenreader would need to know the language to know how to pronounce things

ben_thatmustbeme: as soon as you mix languages, this is always going to be ap roblem
... unless you markup every sentence

sandro: html has 20 years of support for that

<Zakim> KevinMarks, you wanted to discuss html lang and hreflang

KevinMarks: That's what I was going to say. THe microformats response is the lang= on individual elements
... This is existing practce
... The challenge is, you can do mixed languages within a document and I've seen people do that. What we haven't seen so far in microformats is parsing that out and putting it into the parsed json format
... Possibly because we haven't seen much use yet
... Another thing is a Google-imposed cultural bias that assumes people only want one language at once
... They have documented this, they don't recognise multilingual text
... We have ways to express this in html. We should find examples.
... What I haven't seen in html or json-ld is a way of saying here's a phrase that translates multiple ways
... We did have that in OpenSocial
... We had ways of expressing activitystreams stuff with different language markup, but made the same assumption that you're reading in multiple languages

sandro: I just wonder if this is in scope

<hhalpin> Its worth thinking about IMHO

sandro: It's obviously a real world problem, but we're still well behind silos, not sure it makes sense for us to try to get ahead of them on this one

AnnB: I agree with sandro in terms of priority, but it seems like a really important thing especially since we're doing social

<AnnB> I just don't want to lose it

<AnnB> as a concept

Arnaud: don't see anyone else on the queue, we can continue this discussion. Seems like general agreement we should go light on this, may be too early for us to try to tackle this problem
... That's the end of the agenda. Anything else?

sandro: WebEx. THis is the last week with Zakim

<melvster> FYI: jessica was strongly in favour of keeping multi lang

sandro: There will be a phonenumber, or webex software

<hhalpin> In an plus, SIP should work better!

sandro: We'll send an email with the details

<AnnB> yay for Jessica!

sandro: We'll also do a practice next Monday, 24 hours ahead
... And try joining 15 minutes early next week

<AnnB> I definitely don't want to lose the multi lang concept

sandro: IRC will be used the same, webex is just for audio

Arnaud: I've done it with other groups, the biggest loss is the tracking of attendance
... We'll have to have people tell Zakim that they're present on the call
... So the minutes reflect the people attending

<KevinMarks> why webex and not WebRTC?

<AnnB> too early, I think, KevinMarks

<AnnB> I believe they're looking for WebRTC solution on longer term

<KevinMarks> https://talky.io/ is pretty good

<AnnB> right, it's great .. but only handles ~5 people

<ben_thatmustbeme> KevinMarks: can't call in to talky.io can you?

<aaronpk> talky.io isn't good for large groups, beta.talky.io is better at that. also no phone bridge for talky

Arnaud: talks about ways webex might go wrong, peopel should call in early

sandro: I've been using android version, I guess there's an iOS version, that might work well

Arnaud: if you use the client you join the meeting officially

<AnnB> connecting via browser did not work for me with Firefox, but did work with IE

<KevinMarks> I use google voice to phone into the phone bridge here

Arnaud: client offers 3 ways to connect audio

<AnnB> might be some issue with my Boeing-configured FF

<AnnB> dunno

Arnaud: Let us know if you have any questions about webex

<cwebber2> sandro: is there a link on how to clal in via SIP or phone somewhere, or will there be?

<aaronpk> goodbye Zakim! *tear*

<ben_thatmustbeme> good bye Zakim

Arnaud: Thanks for joining, talk to you next week!

<hhalpin> bye bye Zakim :(

Aww, bye Zakim

<Arnaud> trackbot, end meeting

<Loqi> aww, cheer up

<Loqi> aww

<AnnB> waaaa ... Zakim ..

Summary of Action Items

Summary of Resolutions

Approval of Minutes of 9 June 2015 https://www.w3.org/wiki/Socialwg/2015-06-09-minutes

Socialwg/2015-06-16-minutes

Social Web Working Group Teleconference

16 Jun 2015

Attendees

Contents

Approval of last week's minutes June 9th

Tracking of actions and issues

Social API

AS2 language support

Summary of Action Items

Summary of Resolutions