Social Web Working Group Teleconference

08 Sep 2015

Attendees

Present: tantek, csarven, sandro, elf-pavlik, aaronpk, rhiaro, ben_thatmustbeme, tsyesika, cwebber2
Regrets: hhalpin, AnnB, jasnell, evanprodrom
Chair: tantek
Scribe: rhiaro, ben_thatmustbeme

Summary of Action Items

[NEW] ACTION: sandro narrow down possible date ranges for next F2F [recorded in http://www.w3.org/2015/09/08-social-irc]

<trackbot> Date: 08 September 2015

<tantek> scribe: rhiaro

<scribe> scribenick: rhiaro

<elf-pavlik> rhiaro++

<tantek> https://www.w3.org/wiki/Socialwg/2015-08-25-minutes

<Loqi> rhiaro has 155 karma

<csarven> +1

Approval of last weeek's minutes

<elf-pavlik> +1

+1

tantek: Resolved, approval of minutes

<tantek> https://www.w3.org/wiki/Socialwg/2015-10-29

F2F at TPAC

<ben_thatmustbeme> scribenick: ben_thatmustbeme

<elf-pavlik> ben_thatmustbeme++

<Loqi> ben_thatmustbeme has 111 karma

tantek: we discussed the F2F last meeting and were hoping by today we would know if we had enough participants
... looking at the current participants list, I don't see updates from AnnB, or aaronpk
... aaronpk do you have any new info?

aaronpk: its still in progress unfortunately

tantek: anyone else? right now we only have 4 confirmed and rhiaro and csarven are dependant on
... sandro
... unless we have some update by the end of the call we are likely going to have to cancel

<csarven> Sounds reasonable to me.

tantek: does anyone else have anything to update, I'm hoping to be corrected but it looks like only option

<rhiaro> scribenick: rhiaro

<cwebber2> I am more likely able to do that

tantek: We had discussed having a separate face to face maybe in the US, where more people can make it
... Maybe everyone could start thinking about when and where they could do that

<cwebber2> I'll probably be in boston near the end of the month but that's too soon for everyone I'm sure :)

tantek: We would like participation from chairs and everyone who is editing a document, and everyone who is implmeenting
... THese are they key people to have at a meeting

sandro: I could send an email or a doodle poll to figure out..

tantek: If we have enough information to start a doodle poll, go ahead. I don't feel like I have enough. Otherwise send an email.

sandro: What more information?

<elf-pavlik> location first?

tantek: Even a range of dates to choose among?

sandro: We're supposed to give 8 weeks notice, so assuming we pick a week from today that starts us week of Nov 10th
... Presumably we don't want to hold it thanksgiving. That's two weeks before thanksgiving, a week or so after thanksgiving

<ben_thatmustbeme> scribenick: ben_thatmustbeme

<csarven> rhiaro Audio is still okay fo rme.

tantek: lets action sandro to narrow down possible date ranges for next F2F

<tantek> ACTION: sandro narrow down possible date ranges for next F2F [recorded in [[1]|http://www.w3.org/2015/09/08-social-irc]]]

<trackbot> Created ACTION-73 - Narrow down possible date ranges for next f2f [on Sandro Hawke - due 2015-09-15].

<tantek> https://www.w3.org/wiki/Socialwg/2015-09-08#Activity_Streams_2.0

Activity Streams 2.0

<tantek> issue 1: "as:Like (an rdfs:Class) or as:like (an rdf:Property)?"

tantek: we have a couple of issues raised by elf-pavlik, before we deep dive into these on telconf we should make sure we have people effected by these on the telcon or don't waste time on this telcon

<elf-pavlik> https://github.com/w3c-social/social-vocab/wiki/Verbs---owl:Class-vs.-rdf:Property

<elf-pavlik> me :)

<csarven> +q

tantek: are there any implementors here impacted by this issue on the telconf?

<csarven> Sort of..

tantek: if its only elf-pavlik being effected, not sure its best to discuss here
... I looked at the information on this issue, but did not see any link to any implementation actually blocked by the issue. I would recommend that you include links to that issue

<tantek> next issue: "Represent bidirectional connections "

<tantek> https://github.com/jasnell/w3c-socialwg-activitystreams/issues/180

tantek: moving on to the next issue
... I don't understand, this issue says closed
... is there a different issue relating to this elf?

elf-pavlik: no, i don't see this issue as resolved, james just closed the issue

tantek: I think that sounds like a different issue should be pointed to or re-opened.

elf-pavlik: I will propose for next week that we need to figure out workflow for github issues

<cwebber2> I'm not enough in rdf-land to be sure I'm impacted

tantek: same question as before applies, who on the call, is impacted by this issue?
... I think this should be iterated on asyncronously until there are more implementors on the call

scribenick rhiaro

<rhiaro> csarven: first of all, unclear on issues, lots of crossreferencing going on across eamils and github

<rhiaro> ... I think what elf is trying to raise is important for AS2

<scribe> scribenick: ben_thatmustbeme

<elf-pavlik> i'll create new issue for it on gh!

tantek: thats why i was asking if there is any implementors on the call, since James is not here, it doesn't really make sense to discuss
... I think this should be clarified as to what part of the spec actually has an issue, the discussion is more conceptual and less concrete

elf-pavlik: I will try to clarify for next week and hopefully james will be on the call

<csarven> More about the model

<rhiaro> I tried to figure out some of this stuff, but I think definitely better understood by actual implementation. Be good to see examples of options implemented.

tantek: specifically i'd recommend you try to get much clearer on what specific parts of the spec are effected

elf-pavlik: I wanted to clarify the approach of asking for implementors on the call since only james implements a library. I think we can discuss the approach before we implement
... specifically this deals with how we model, i don't think we need to have implementors on the call

tantek: they could be implementING, don't have to have implmentED. so its who feels this would effect them

<tantek> https://www.w3.org/wiki/Socialwg/2015-09-08#Social_API

moving on

Social API

<tantek> "User Stories -- what are the implications of a story being "accepted"? "

tantek: question was raised as to what are the implications of a story being "accepted"
... amy are you back on to take over as scribe?

I'm willing to let elf-pavlik introduce

<melvster1> FYI: there 519 +1 votes -- "will implement"

elf-pavlik: we have had back and forth on what are the consequences of accepting a user story

tantek: that should be documented at the social api page

<rhiaro> scribenick: rhiaro

tantek: We were talking about what should be in the social api. We had research on what proprietary social APIs do, all on the wiki
... Evan then proposed a list of social api requirements, based on a union of all those proprietary APIs
... The group decided that was far beyond what we needed to do for a v1

<elf-pavlik> link to minutes?

tantek: Instead of using a list of derived reqs from existing sites like that, we wanted to base the requirements on user stories that the group was interested in solving
... That led to the proposal of user stories, and voting on them
... After that ben_thatmustbeme sorted them by different clusters

<tantek> https://www.w3.org/wiki/Socialwg/Social_API/Sorting_user_stories

<ben_thatmustbeme> elf-pavlik, i have all the minutes downloaded, i can grep across them for that week to link you to probably

<ben_thatmustbeme> after the meeting that is

tantek: After that, or even before that, we agreed that swat0 was one of our foundational user stories
... Then we agreed to accept the user stories that were all +1

<ben_thatmustbeme> not at this computer

tantek: Leaving the rest to the IG to attempt to resolve these issues on
... The implication there being that the purpose of the user stories is to drive rquireemnts for the API, so the accepted user stories would be used to dtermine those requirements
... Maybe that's the next step, to say given these user stories what does that say about reqs for API
... That's how we got to where we are

<rhiaro> I tried the reqs from user stories doc

sandro: We should avoid being legalistic about it. Anyone designing an api should have in mind the things that everyone seems to want it to do

<melvster1> sandro++

sandro: Not an agreement that we will only consider things that will only meet all of these criteria, and have some kind of formal thing

<Loqi> sandro has 23 karma

<ben_thatmustbeme> we should probably be written on the site, or at least linked anywhere there is a list of "approved" user stories

sandro: Makes people panic and argue in a way that doesn't reflect reality
... We should keep it more descriptive than prescriptive
... But that's my perspective, not always how people want to do things

tantek: I think your perspective is accurate

<ben_thatmustbeme> sandro++

<Loqi> sandro has 24 karma

<elf-pavlik> sandro++

<Loqi> sandro has 25 karma

tantek: I think the user stories list and voting was an attempt to narrow absolute list and requirements
... These are good inputs to what a social api SHOULD have rather than MUST have is a good interpretation
... That's my understanding as to how we got here

elf-pavlik: I get frustrated that you and sandro see differently. You said it's to narrow the absolute requirement, and sandro said it's more about getting general idea of what people want, not absolute
... So we don't look for this absolute requirement or hard line, but what we can expect others to give attention to, and maybe we should focus furst on things that are of common interest, and not feel discouraged about pursuing things that are less of common interest
... We should document this, and why do we have a formal process of accepting and not accepting, seems we have a different understanding

tantek: your point about prioritising what seems to be common wants is a good one
... Certainly the tone in the group has shifted from the earlier days when we were trying to require all of these things
... Now people are implementing it's clear that there's a lot of work to be done
... Totally fine to move in that direction. What sandro described makes sense, without trying to compel what goes into a draft
... The more we canminimise the MUSTs the better

<hhalpin> +1 should, be we need interop over a small set of minimal functionality

tantek: Makes it easier to write a spec, test suite, go to CR

<hhalpin> for the CR test-suite

elf-pavlike: Maybe we don't need this process for accepting user stories?

tantek: We need to know what anyone wants

<melvster1> people are not implementing ... there were 519 +1s meaning "will implement" ... but only a small fraction of those voters followed through ...

sandro: I lean against a formal process. We can base it on observation of votes. People can change their votes if priorities change

tantek: Does anyone know what melvin means by 519 +1s?
... I don't udnerstand what that's about

<melvster1> the user story voting, I counted the +1s

ben_thatmustbeme: I'm guessing he jsut counted all the +1s between all the votes?
... Because it was originally that +1 was labelled as 'plan to implement'

tantek: THat's a misinterpretation
... just because someone +1ed doesn't mean they were planning to implement

<melvster1> ok, it's on the page still

tantek: Some people noted they would implement, but +1 doesn't mean that
... Going to treat that as misreading documentation

<aaronpk> melvster1, what are you using to determine whether someone is implementing? I am implementing things and still plan to implement things that I have not started yet, so not really sure what point you're trying to make

<rhiaro> For the record.. https://www.w3.org/wiki/Socialwg/Social_API/Requirements

<melvster1> +1 "yes, I need it, will implement, worth doing"

tantek: Whether user stories are useful to accept or not, it's to help understand what's common interests

<melvster1> http://www.w3.org/wiki/Socialwg/Social_API/User_stories

tantek: Just because we're moving away from MUST to SHOULD doesn't mean we throw everything out
... If you can't describe some api functionality in terms of user story, probably means you shouldn't have it
... To stop draft ballooning out of control with every feature request

<ben_thatmustbeme> melvster1, probably you should read those ,'s as ORs

<ben_thatmustbeme> not ANDs

elf-pavlik: Will we continue process of trying to formally +1 and accept? Or keep it around as a reference but no more formalising?

tantek: Depends on what desires the different impelentors have?
... Are there user stories that people are implementing that go beyond the scope of what is already approved or not?
... As I hear, even the user storeis that are accepted are a lot of work, so we don't need more
... I strongly suggest that if you're implementing something you focus on what's already been approved, instead of trying to add more work
... When someone is willing to declare victory on implementing all the accepted user stories, we can have this discussion again

<rhiaro> melvster1, I read the ',' in "yes, I need..." as OR rather than AND

tantek: ben, anything to add about implications of user stories?

ben_thatmustbeme: I guess not, mostly covered
... A lot of this came out of people trying to implement those exactly. We need to make it known that people don't need to implement every exact thing. Not specs, just reference for guiding us.

tantek: What do people think about that? Implementors should take the user stories as useful functionality, but not a precise step by step requirement

ben_thatmustbeme: exactly

sandro: I guess I wonder why that is?
... If you ahve 3 different implementors demoing user stories, and two do it one way and one is different, that seems like it's going to cause confusion and problems

ben_thatmustbeme: The problem is that we didn't read these user stories as exact specifications. If I went back and voted on them that way it woudl be different. We ended up arguing about 'inbox' terminology

sandro: Terminology doesn't affect code, so terminology aside, what else?

tantek: I don't think we can avoid people misinterpreting a user stories

ben_thatmustbeme: I didn't go through these thinking of them as specs. The functionality of being able to like something is different to having a button on a site below a post
... If we start reading them as technical documents, you have to follow exactly what they say
... It's important to tell people that's not what these are, because they were not gone over that way
... We should not say this is approved, this is perfect, you should implement exactly like this
... that may not make sense

sandro: If we say that these are not to be understood precisely, we've muddied the waters without gving any new guidence

tantek: not if that's hwo we all interpret them
... If the folks that voted on them initially did not interpret them with thta precision, it does not seem reasonable to assert that precision now

<hhalpin> I imagine the 'precision' will come out in the test-suite, right?

sandro: I agree with that, just trying to find something more useful than handwaving around it
... My proposal would be that anyone that thinks this goes back to change their plus ones to clarify this

<hhalpin> It seems this 'terminology' issue might be making a mountain out of a molehill.

sandro: I'm fine with saying terminololy everywhere is not rigid
... But other things. Seems more useful than just saying don't take any of this stuff literally

tantek: Let's at least put that proposal... let the user stories be interpreted as stheir functionality, not their terminology

sandro: I'm cool with that

tantek: Does that reflect at least part of your concern, ben?

ben_thatmustbeme: certainly

tantek: any objections?
... if you have objections, -1 in IRC
... This is a specific proposal
... Then we'll go to the queue
... Let's at least put that proposal... let the user stories be interpreted as stheir functionality, not their terminology

<tantek> +1

<csarven> +1

<ben_thatmustbeme> +1

<cwebber2> +1

<cwebber2> seems good

PROPOSAL: User stories are interpreted as their functionality, not by terminology

<elf-pavlik> +1

+1

<tsyesika> +1

sandro: +1

RESOLUTION: User stories are interpreted as their functionality, not by terminology

tantek: Hopefully thsi will resolve literal interpretation problems in future
... Next issue is, inbox user story was overinterpreted
... I don't think we can avoid that, people are going to interpret things however they want, people are looking for guidence wherever they can get it
... Okay if osmeone wants to literally implement a user story, but not okay to demand others do so
... I believe saying the functionality is the focus makes that clear
... To avoid these arguements by literal text in the future

ben_thatmustbeme: I think that'll take care of it

<Zakim> elf-pavlik, you wanted to discuss IG ambition to 'clarify' those stories and need for it

elf-pavlik: I would like to see if we give feedback to the interest group, who I hope will get more active now
... They need to know if they should put effort into resolving issues to get them accepted
... If we want to continue process of formal approval and have IG put effort into clarifying those stories so they can be formally approved, we still need to write down clear criteria of what acceptance implies
... If it's just for understanding each other, we don't need all this strong effort into formal acceptance process, with people putting effort from IG into this

tantek: THere have been several times people have stated that it's useful to have a set of user stories that are accepted to prioritise common interests
... Many people have indicated why that's useful
... As far as trying to provide criteria to the IG, I don't see why that's necessary in terms of what the user stories are used for
... The request of the IG is to go through 0s and -1s on user stories and resolve them
... That is to contact people objecting or being neutral about a user story, finding out why, seeing if it's possibe to resolve, ideally with original author of user story
... It doesn't need anything more than that. Focus on those votes and those people, resolve them one by one.
... Anyone here from the IG to take that action?
... AnnB is not here
... Anyone an active participant in IG to take that action?

<rhiaro> When the IG did try to do thsi a while back, WG members didn't respond well

tantek: Not hearing anyone step forward to take this action to the IG, so going to postpone. elf, you can try bringing this up again in the future. Unless we get IG participatns in the WG, we're not going to get progress on this
... If you want to drive that, you need to drive it in the IG yourself

ben_thatmustbeme: Going to suggest that maybe the idea of having a formal approve/deny might be easier to do just as once we all agree on it, everyone is +1, then it's implicitly approved

<elf-pavlik> +1 ben_thatmustbeme

ben_thatmustbeme: This takes away any work on us to waste conference call time to put forward to proposals to approve
... If everyone agrees they see it as useful, it's implicitly accepted

tantek: appreciate the optimism. If those vote changes occur based on changes to user stories, I think it's reasonable to have to raise those user stories to everyone else who might have voted, to make sure that their votes don't change
... So we're not specifically asking everyone to go vote again, just trying to resolve non +1 votes with edits
... With the optimistic assumption that everyone who voted +1 won't change their vote
... But we have to challenge that assumption by bringing it up in a telecon to check for objections
... Don't want to edit user stories from underneath people without them being notified

<ben_thatmustbeme> makes sense

<tantek> https://www.w3.org/wiki/Socialwg/2015-09-08#Tracking_of_Actions_and_Issues

Tracking of actions and issues

<tantek> http://www.w3.org/Social/track/issues/raised - none

tantek: 0 raised issues

<tantek> http://www.w3.org/Social/track/issues/pendingreview - none

tantek: 0 pending review issues
... 0 pending review actions either

<tantek> http://www.w3.org/Social/track/actions/pendingreview - none

tantek: Next telecon is next week
... 15th, Arnaud chairing
... Any other issues that anyone wants to raise in last 2 minutes?
... We're cancelling the face to face at tpac unfortunately as people can't make it
... We'll follow up with a doodle poll to figure out when we can meet next

<elf-pavlik> tantek++

tantek: Thanks everyone

Socialwg/2015-09-08-minutes