- DRAFT -
Social Web Working Group Teleconference
03 Feb 2015
See also: IRC log
- Ann, Arnaud, aaronpk, tantek, jasnell, AdamB, elf-pavlik, bret, rhiaro_, eprodrom, bblfish, dromasca, Lloyd_Fassett, bill-looby, Sandro, hhalpin, Tsyesika, +1.408.335.aaee, +358.503.28aaff
<trackbot> Date: 03 February 2015
<tantek> scribenick: aaronpk
<scribe> chair: tantek
tantek: first order of business is to approve last weeks minutes
<bill-looby> zakim ipcaller is me
<eprodrom> tantek I believe I have an open action to add the IBM Connections discussion to the agenda
okay then up to evan toa dd the items to the agenda before the end
<AdamB> looked very lively :)
any opinions on last weeks' minutes?
scribe: shall we approve them?
scribe: not hearing any objections
... based on no objections and one +1, declare the minutes approved
... next item
tantek: first issue, should we drop verbs and only use object types
scribe: any suggestions for how we resolve this since it's not assigned to anyone
<jasnell> this has already been done in the current working draft
Arnaud: we can give someone an action to come up with a proposal to address an issue
<Zakim> elf-pavlik, you wanted to discuss ACTION-26 Review microformats due jan 14, 2015
<elf-pavlik> i'll type
<bret> (sound is breaking up... sounds like a mic level issue)
<bblfish> can't hear Elf either
tantek: if this is about the open issue type in IRC, otherwise will ack james
jasnell: re: verbs, the current draft already dropped verb
<Loqi> Tsyesika: elf-pavlik left you a message 3 weeks, 6 days ago: sorry for ping just checking Loqi ...
jasnell: the verb property from the original AS has already been deprecated
<eprodrom> +1 resolved
<elf-pavlik> +1 resolved
<eprodrom> I just closed it
tantek: next issue explicit or implicit typing
jasnell: the current approach is to depend on explicit types. the implicit approach was talked about at the F2F but not followed up on
... i'm sure an implementation could choose to do implicit if it wanted, but right now the approach is explicit
tantek: is there anyone here who would like to propose implicit typing?
<bblfish> I suppose implicit typing would require rdf:domain and rdf:range
tantek: specifically how to do it
<eprodrom> Not me and I believe the current typing is good
bblfish: I don't have an opinion on this, but if you wanted to, you'd specify the domain and range
... and then an rdf inferencing engine would be able to infer it
... but we don't want to make people rely on this in the beginning, better to make it explicit
<hhalpin> yep, its highly unlikely anyone will use RDF(S) inferencing.
<elf-pavlik> tantek, i think you came up with it so maybe you could take action to research it further?
tantek: are you proposing a way to do it, or are you saying here are some thoughts
bblfish: i'm fine with explicit typing, if youwant implicit you have to do domain and range
<bblfish> If I am wrong then it's because I have not understood the problem
tantek: I was one of the ones who brought this up at the F2F so I'l ltake the action to come up with a simple proposal for implicit typying
... based on property names
<eprodrom> Create an action
<scribe> ACTION: tantek to come up with a simple proposal for implicit typing based on property names https://www.w3.org/Social/track/issues/4 [recorded in [|http://www.w3.org/2015/02/03-social-minutes.html#action01]]]
<trackbot> Created ACTION-35 - Come up with a simple proposal for implicit typing based on property names https://www.w3.org/social/track/issues/4 [on Tantek Çelik - due 2015-02-10].
tantek: next issue: need glossary for terms https://www.w3.org/Social/track/issues/5
... ann did you raise this at the f2f?
... we were using a lot of jargon at the f2f and we didn't all agree necessarily
<bblfish> I wonder if this can be subsumed by use cases
AnnB: we do have lloyd, chair of the vocab task force in the IG
tantek: no this isn't for vocabulary, just in prose
AnnB: a glossary is similar to teh vocabulary no?
eprodrom: the problem was that we have a use case that speaks abotu federating profiles, uses "federate" as a verb without clearly saying what that means
... the idae was that we would grab some of the jargon terms and nail them down
... i'm happy with closing the issue, don't believe it's necessary. i'm happy to live with the ambiguity
<AdamB> +1 on waiting until another problem
tantek: anyone else happy to live with the ambiguity until we run into a problem?
AnnB: if we encounter a specific issue we can address then
tantek: proposeal to closing with no resolution
<eprodrom> I closed it
lloyd: my perspective is federation through vocabulary is a circular argument
... to me it was a discussion about having a central resource to do the federation or is the standard going to allow direct multi-point to multi point federation
tantek: that's not what the issue was about, just that we couldn't agree on what federation meant
<eprodrom> hhalpin: thanks
lloyd: agree with living with the ambiguity for now
<hhalpin> IMHO this "federation" conversation will go on forever
bblfish: suggestion to drop "federation" and use "p2p"
<eprodrom> tantek can we just say "any updates to actions and issues" and move on?
<hhalpin> so let's take that to the IG mailing list
tantek: sandro can I assign this to you?
<tantek> ISSUE-6: Make sure we have teleconferencing equipment for next f2f
<trackbot> Notes added to ISSUE-6 Make sure we have teleconferencing equipment for next f2f.
<scribe> ACTION: sandro resolve issue 6 https://www.w3.org/Social/track/issues/6 [recorded in [|http://www.w3.org/2015/02/03-social-minutes.html#action02]]]
<trackbot> Created ACTION-36 - Resolve issue 6 https://www.w3.org/social/track/issues/6 [on Sandro Hawke - due 2015-02-10].
tantek: next issue, 8
<tantek> ISSUE-8: Test suite for activity streams 2.0
<trackbot> Notes added to ISSUE-8 Test suite for activity streams 2.0.
<trackbot> ISSUE-8 -- Test suite for activity streams 2.0 -- open
<bill-looby> Are we discussing Issue 7 ? or does pending review mean addressing outside this call ?
tantek: we don't currently have any concrete plans for a test suite
... anyone want to volunteer to work on at least a plan for a test suite
bblfish: recently looked at a case of an ontology written out, and these are difficult to test currently
... so what we might want to do is ask the people from OWL or RDF working groups to find what tools they use to test ontologies
tantek: hold on we don't need a specific plan right now, are you willing to agree to take the action
bblfish: i'm just saying why don't we ask other people what tools we use
<bblfish> I am saying why don't we ask people in the OWL / RDF land
sandro: having been in those WGs there's nothing useful
<scribe> ACTION: hhalpin put forth the test suite plan using standard JS tools https://www.w3.org/Social/track/issues/8 [recorded in [|http://www.w3.org/2015/02/03-social-minutes.html#action03]]]
<trackbot> Created ACTION-37 - Put forth the test suite plan using standard js tools https://www.w3.org/social/track/issues/8 [on Harry Halpin - due 2015-02-10].
tantek: next issues 9 and 10
... skip those for now since that's what we're working on
... any actions that anyone wants to report progress on?
tantek: if nothing to report, we'll move on to the rest of the agenda
<harry> i'm off queue
jasnell: the AS drafts were published on thursday
<Loqi> I added a countdown for 2/5 12:00am (#5603)
<eprodrom> aaronpk: yes
bill-looby: been looking at proposed API, but still don't have access to the wiki
<AnnB> ++ to jasnell
tantek: okay next agenda item, move forward with work on the social api. evan's action so he can lead the discussion
eprodrom: thanks tantek. so last week we had an action item to approve a list of requirements
... which opened up a lot of conversations
... which is great in a lot of ways, but is also difficult cause we can go all over the place
... had a lot of discussion on the mailing list and on the telcon last week
... one of the big items of feedback is the format of the requirements we have is probably not sufficient to develop a candicate proposal for
... in discussion with teh other chairs, in order to move forward, we proposed a mechanism and schedule for pushing the process forward
eprodrom: the idea is in three steps
... first step is take the list of requirements we have now,
... things like creating social connections, posting new content, repsonding to content
... using each fo the high level groupings to create a user story
... we have about 10 high level groupings, so we create a set of 10 user stories
... would have those ready to go for next week, Feb 10
... and then if we generate more user stories in that time that's great, btu can definitely commit to taking the requirements we have and turning them into user stories
... will be doing this on the wiki so people can review
... we'l talk about it next week Feb 10, and over the week of 10-17 working group participants will use +1 -1 +0 voting on the wiki page
... to upvote/downvote user stories
eprodrom: on 17 feb we come out with a set of 0-10 user stories that we use to measure our candidate proposals
... there's nothing in this keeping us from developing candidate proposals
... from the point of view of the chairs and w3c team, if we have a good set of user stories we agree on, it will be much easier t oaccept the validitiy of proposals that come in
<KevinMarks> if we have 0 user stories does that mean we have no specs to do?
eprodrom: any questions abotu the procesS?
tantek: would like to point out there is one user story to be used as at template for developing user stories
eprodrom: we have one user story previously approved, SWAT0, a good example of a concrete user story that is a good model for other user storie that come in
<bblfish> I'd say "very simple"
<eprodrom> bblfish: thanks
AnnB: i feel like the IG failed the WG in that we were supposed to define a bunch of use cases but it didn't happen
<AdamB> how does the swat0 use case differ from something like this: https://www.w3.org/wiki/Socialig/Use_Case_TF/Profile_Use_Cases/Social_Profile_Creation
AnnB: adam and I have spent some time int he last couple days, we have a ton of use cases described in a narrative format
... but we didn't get them into a common use case template
... after this meeting last week when there was discussion about havsing distinct step 1-2-3 use cases, tried to do that with one of them and adam tried to do a few more
... my question is your call now for user stories, is that comparable to use cases?
... what more oculd we do here to be useful
eprodrom: i think the idea behind user stories is 3-5 sentence paragraph that talks about a process someone goes through satisfying a need
... there may be multiple functional points in that process, but we're talking abotu end users satisfying requirements using a particular tool
... also does anyone have an understanding of user stories that is far different fomr that?
tantek: there is a key implicit question of "why"
... it's good to document in steps, ann's example of social profile creation is a good step in that direction
... when evan and i developed swat0 years ago, there was a clear motivation of why behind the user story
... this is osmething people do today, they take pictures and post them and tag each other
... it was based on existing behavior
... the why is an important aspect of any use case
... the example ann posted of social profile creation, people are creating profiles all the time, its' useful to ask why are they doing that
... there are no service where you just createa profile for its own sake, you're doing it for a reason
AnnB: i was thinking the narrative in these describes the why
... in the case of profile, pretty much all other actions depend on the profile being createdin the first place
... were trying to boil it down into twitter-like succinct steps
... we have some exmaples out there, would be useful to get some feedback now
... would be interested to know how these are different from user stories
<tantek> note that last week there was consensus about requiring "tweetable" summaries of use-cases / user-stories, so I summarized as such for SWAT0: http://tantek.com/2015/029/t1/swat0-posts-tags-mobile-photo-comment and in tweet form: https://twitter.com/t/status/560921112767242241
tantek: since you bvrought up the point about having tweetable user stories
<Loqi> @t :: SWAT0
<Loqi> A posts+tags mobile photo of B
<Loqi> B photo notified
<Loqi> C(follows A) sees it; replies
<Loqi> A&B comment notified
tantek: would like to put this forth as a proposal
PROPOSAL: any user stories that are posted have a tweetable basic summary of the steps
<KevinMarks> can we use images to game the tweet limit?
AnnB: might be too brief. in some cases we might get stuck tyring to make it too short
<sandro> KevinMarks, you mean like a screenshot of twitter rejecting your really long tweet? :-)
tantek: the point is if it feels like it's not fittin,g it probably needs to be broken into multiple stories
AnnB: the IG has focused ont he profile use cases because pretty much everything derives from a profile
... but perhaps we should expand our horizons now
sandro: mostly want to ask evan about this, thought i heard him volunteering to rethink the requiremetns into a corresponding set of use cases
... while I love the precision of the profile use case and swat0, but even sketching out briefly to find out if that's what people want out of the WG
<AnnB> seems to me there needs to be more than SWAT0
sandro: once we have that we can delve more into fleshing them out
<AnnB> but which?
<AnnB> and how do they get approved?
<tantek> welcome Lloyd_Fassett
<tantek> (on IRC)
AnnB: i'm now chairing the IG, trying to focus that energy into whatever is useful in the WG
... that's our mission is to be supportive of teh WG
eprodrom: we could probably talk about the purpose of the IG
... i would love to take the requirement list and convert them into user stories, waill try to keep them short
... may be an interesting exercise to make them tweetable
<bblfish> Frankly Tweets are a bit overrated
eprodrom: the help of the IG over the next week would be to review those
<KevinMarks> use emoji instead of A,B,C : https://twitter.com/kevinmarks/status/562682597772627968
<Loqi> @kevinmarks :: "@t: SWAT0
<Loqi> 👱posts+tags mobile photo of 👲
eprodrom: this is really about getting the social API out
<Loqi> 👲photo notified
<Loqi> 👳(follows 👱) sees it; replies
<Loqi> 👱&👲comment notified
<Loqi> Prev: http://t.co/U3j2gqf36Q"
<Lloyd_Fassett> thanks tantek!
eprodrom: there's a federation case in the profile use case, but I don't think federation is applicable to social API so not somtehing we'll deal with this time around
<tantek> bblfish: the "tweet" constraint is a deliberate forcing function for simplification, and reducing "essays" which tend to otherwise occur
eprodrom: effort is to get the drafts up ASAP and have the IG review them
<AnnB> good feedback, eprodrom; thanks
<tantek> make that bblfish, the "tweet" constraint is a deliberate forcing function for simplification, and reducing "essays" which tend to otherwise occur
eprodrom: I might even be able to have them ready for tomororw's IG call
<scribe> ACTION: eprodrom to convert social API requirements to 2-10 user stories of 3-5 sentences each [recorded in [|http://www.w3.org/2015/02/03-social-minutes.html#action04]]]
<trackbot> Created ACTION-38 - Convert social api requirements to 2-10 user stories of 3-5 sentences each [on Evan Prodromou - due 2015-02-10].
tantek: beyond that the general proposal for user stories to drive requirements by tuesday, then to evaluate the user stories and vote by the 17th
<bblfish> will there be enough time for people to add their own stories, in case they find some missing?
tantek: that voting should include: +1 you would use that yourself, but also +1 you would implement that in a client or +1 you would implement that on a server
<sandro> +1 voting should be about whether you'll implement on server, implement on client, buy products using it, ... etc, or somehow see it as harmful or problematic.
tantek: you have a week to propose a user story, then you have a week to vote on whether you would *use* and *implement* those stories
... or if you have already implmeemntned it
<AnnB> note that some of us (e.g., me) don't implement things
<sandro> AnnB, you can speak to the products that Boeing wants
tantek: anyone is welcome to take any of the work done in the IG and put it forth for the WG to consider
... this is a general call for anyone in this WG to take any user story (they don't need to write it themselves) and prppose it
hhalpin: quick note, effectively we're doing this process because we had such disagreement over the requirements
... this seems like the best way forward
... we don't have to accept all the user storeies
... what we didn't want is open-ended user stories
... what we wanted is user stories that we use to psuh out the API draft and know the benchmarks of it
... at least a rough consensus
<AnnB> right, sandro .. my point is I don't want something ranked lower because I don't vote that I will implement it
<jaakko> my team would like to be building a prototype in the upcoming weeks but we're still uncertain about the technology choices. we have the user story and source data to be consumed and published in a activity stream. what would be the best way to go about building a proto?
hhalpin: we did rat-hole really badly last week, we had users bringing up implementation level details abotu how they wanted the plumbing to work
... we dont' want to do that in the user stories
... they have to be implementation independent at this point
... let's try to stay on focus on the mailing list, and if people want to have more technical discussions about plumbing, that's off topic for the WG and you can do that in the IG
eprodrom: henry it soundsl ike you're pretty eager to do user stories, you can drop them in on this blank page
<harry> Again, we had to do this because we didn't have agreement on requirement list
eprodrom: it was my intention to get candidate stories up ASAP, if you have something you'd like to throw in there you can do it
... I did not mean t osuggest it was not ok to add user stories
<harry> So rough consensus on a small set of user-stories seems reasonable to push through
tantek: okay does that help clarify the proposal?
... does everyone understand this week vs next week?
<harry> Yes, chairs have regular meetings, that's normal AnnB
<eprodrom> Text I had was " PROPOSED by 10 Feb user stories on wiki to be voted on between 10-17 Feb with approval on 17 Feb"
<bblfish> that's fine, just trying to get the understanding of the timetable clear.
<bblfish> When can we expect the first stories to be down?
<harry> particularly when Working Group needs to focus on getting chartered deliverables though.
<eprodrom> bblfish I can't commit to having them ready before 10 Feb
<AnnB> +1 (I think)
<sandro> +1 with "votes" being feedback about user stories, and who cares about them, more than up-vs-down
<eprodrom> But I will try to work on them in the next 48 hours
<bblfish> +1 though we need to have some template stories up a.s.a.p so that we know how people want to write up u.stories
<tantek> +1 for "I would use this user story personally", +1 would implement this story on the client, +1 for would implement on the server
<AnnB> oh .. that's not what I thought this vote was about
sandro: we'l see a few user stories come out as most exciting
tantek: important to capture implementer interest as well
<eprodrom> AnnB: they're describing what the "voting" on the wiki will mean
<eprodrom> It's kind of recursive, sorry
tantek: not seeing any objections to the schedule
<AnnB> I thought this particular vote was about the schedule for next 2 weeks
<eprodrom> AnnB: yes, correct
RESOLUTION: accept evan's proposal for the next two weeks. week 1 is to propose user stories on the wiki, week 2 is to vote on them
bblfish: evan just wrote that he won't have anything up by the 10th
tantek: you have until the 10th to add to the wiki page
... from the 10th to the 17th you have the opportunity to vote
<KevinMarks> which is it?
bblfish: is there going to be some discussion about.... it seems like you're trying to push things through at full speed
tantek: what's been happening for months is we've been reviewing existing APIs
... that research has been used to inform the draft list of requiements
... which was met with lots of discussion and dissent last week
... evan's proposed way of moving past that is to ground the requirements in user stories we can use to drive the requirements
bblfish: worried that people will be writing user stories in the wrong format, and a week isn't enough time
... have to be clear on the type of the story, feeling this is a bit pressed
tantek: this question has been answered, we have one great story, if you have any question about the format, use that
<AdamB> swat0 uses a numbered list for "user story"
tantek: any questions bring it up on the mailing list
bblfish: i dont' think you're givingp eople a lot of time to iterate here
tantek: there has been a lot of iteration in the IG
... we just didn't provide any structured way to bring the stories forward, so this is the structure
<AnnB> BUT, my point is, the IG has tried to put use cases forward in the same format as SWAT0 .. what next?
<jasnell> Propose one additional week
<jasnell> two week review
<KevinMarks> we have a call netx week to review, and another week, and we can still iterate
<eprodrom> Sorry, I dropped off
hhalpin: want to reinforce. "hi I want this whole thing to use RDF" is not a user story
... and B we cannot go out of scope
... things like identity systems are out of scope
... we cannot and should not deal with out of scope things here
... in the mailing list we had a lot of rat-holing, which is on topic on a different list
<hhalpin> AnnB, that obviously didn't work though, which is why we are trying to reduce the amount of text required.
eprodrom: we are trying to rush, i tehink ideally speaking for the chairs, we would love to have proposals on the table and be able to select a proposal at our F2F
... feeling like we can move forward with the social API at the f2f
<AnnB> did you look, hhalpin, at the ones we just wrote?
<Arnaud> bblfish: while I can understand you think we are pushing that's what chairs have to do, everyone will have the opportunity to comment, ask questions, etc. so I don't think you need to worry
eprodrom: if we take this extra time that you are asking for that you put it t ogood use
... i'm willing to put in some extra time to get these early user storeies as fast as possible
<hhalpin> Not the new ones, and I'd just add them immediately to the wiki if possible, as that won't take too much time.
eprodrom: i don't think it will take too much editorial work
... i would like if you're putting this extra burden on me that you make good use of it
<KevinMarks> instead of extra time now, propose an extension next week if needed then
<AnnB> they are there, Harry
bblfish: if you can get as many of them done in 2 days time, then other people can look at them and discuss and add to them
<KevinMarks> we can always iterate furthre
bblfish: because you already have a good idea of the use caseas you want
... just so we have a better idea of what you're looking for
<AdamB> i've taken a shot at adding one https://www.w3.org/wiki/Socialwg/Social_API/User_stories#Proposed_User_Stories
bblfish: will be much easier if we have 4-5 days after seeing yours to add to them
<AnnB> Profiles: https://www.w3.org/wiki/Socialig/Use_Case_TF/Profile_Use_Cases
tantek: sounds like we have a consensus
<tantek> schedule resolution stands
<AnnB> just before the meeting, I moved the one about Federated Groups to another page; didn't have to do with profiles
sandro: the important thing is that we have solid consensus ont he user stories
... hopefully everyone agrees in 2 weeks
... but if in 2 weeks some people ahven't read them
... we should try our best to actually get feedback from the overwhelming majority of members
<eprodrom> OK, dropped off
tantek: thanks everyone
Summary of Action Items
[NEW] ACTION: eprodrom to convert social API requirements to 2-10 user stories of 3-5 sentences each [recorded in http://www.w3.org/2015/02/03-social-minutes.html#action04]
[NEW] ACTION: hhalpin put forth the test suite plan using standard JS tools https://www.w3.org/Social/track/issues/8 [recorded in http://www.w3.org/2015/02/03-social-minutes.html#action03]
[NEW] ACTION: sandro resolve issue 6 https://www.w3.org/Social/track/issues/6 [recorded in http://www.w3.org/2015/02/03-social-minutes.html#action02]
[NEW] ACTION: tantek to come up with a simple proposal for implicit typing based on property names https://www.w3.org/Social/track/issues/4 [recorded in http://www.w3.org/2015/02/03-social-minutes.html#action01]
[End of minutes]
Scribe.perl diagnostic output
[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.140 of Date: 2014-11-06 18:16:30 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/???/bblfish/ Succeeded: s/hhalpin/sandro/ Succeeded: s/???/sandro/ Found ScribeNick: aaronpk Inferring Scribes: aaronpk WARNING: No "Present: ... " found! Possibly Present: AdamB Ann AnnB Arnaud IPcaller ISSUE-6 ISSUE-8 JakeHart KevinMarks Lloyd_Fassett Loqi Mozilla P16 P18 P5 P8 P9 PROPOSAL Prev Profiles ShaneHudson Tsyesika aaaa aabb aacc aadd aaee aaronpk almereyda almereyda_ bblfish ben_thatmust ben_thatmustbeme bigbluehat bill-looby bret danbri dromasca dwhly elf-pavlik eprodrom harry hhalpin https jaakko jaakko_ jaensen jasnell jaywink joined kylewm lloyd mattl melvster nickstenn oshepherd_ pdurbin rektide rhiaro rhiaro_ sandro scribenick shepazu social tantek the_frey trackbot wilkie wseltzer You can indicate people for the Present list like this: <dbooth> Present: dbooth jonathan mary <dbooth> Present+ amy Agenda: https://www.w3.org/wiki/Socialwg/2015-02-03 Found Date: 03 Feb 2015 Guessing minutes URL: http://www.w3.org/2015/02/03-social-minutes.html People with action items: eprodrom hhalpin sandro tantek