Socialwg/2016-02-16-minutes

From W3C Wiki

16 Feb 2016

Agenda

See also: IRC log

Attendees

Present
aaronpk, Rob_Sanderson, Arnaud, ben_thatmustbeme, sandro, rhiaro, AnnB, KevinMarks, dmitriz, cwebber2, dromasca, tsyesika, Natalie
Regrets
eprodrom, tantek, jasnell
Chair
Arnaud
Scribe
rhiaro

Contents





<scribe> scribenick: rhiaro

Approval of minutes

Arnaud: two sets to approve, last week, and from a copule of weeks ago

<ben_thatmustbeme> yeah, i think they were fixed right after last meeting

<Arnaud> https://www.w3.org/wiki/Socialwg/2016-02-16

<Arnaud> Approval of Minutes of 2016-01-26

<Arnaud> https://www.w3.org/wiki/Socialwg/2016-01-26-minutes

Arnaud: Any objections?

+1

<azaroth> +1

<ben_thatmustbeme> +1

<cwebber2-2> +1

RESOLUTION: Approved Minutes of 2016-01-26

Arnaud: And minutes from Feb 9

<Arnaud> Approval of Minutes of 2016-02-09

<AnnB> +1

<Arnaud> https://www.w3.org/wiki/Socialwg/2016-02-09-minutes

<ben_thatmustbeme> +1

<dromasca> +1

<cwebber2-2> +1

RESOLUTION: Approved Minutes of 2016-02-09

Arnaud: Unfortunately when I was trying to put the agenda together yesterday it seemed not much had happened
... I carried over topics from last week, mostly about AS2 publication
... Trying to get there, no open issues, mostly a matter of getting the draft in shape so it can be published and reviewed by the group so that we can move on to CR
... jasnell has been super busy with another project
... every week I ask him what the status is and he tells me that he's going to get it done this week, said the same thing yesterday, I did point out he tells me that every week, but hopefully he will eventually get to it

sandro: I'm surprised to hear you say that there are no open issues; what happened to the issues we agreed to open a couple of weeks ago like conformance clause

Arnaud: I meant no technical issues
... There's still the issue of conformance and definition of the features
... I didn't expect Evan to also be absent today, he took the actions to take a crack at those things
... The agreement is that he would take a crack at addressing those things and we would publish the new draft and the group woudl review and discuss, and assuming this gets settled we'd be able to close this
... Then the only thing left would be the test suite

Introducing Natalie from Thoughtworks

sandro: I just realised that we have a guest on the call who I was going to introduce
... Natalie, can you hear me?

Natalie: Hello

sandro: You all my recall Harry Halpin was representin gthe dcent prorject which was an EU research project on decentralised technologies, and there were industrial partners implementing some of our technologies
... Natalie is a representative of one of those, so we asked if she could join the call to give their feedback on AS2
... With James and Evan not here, there might not be much use today

<ben_thatmustbeme> Still scribing

sandro: But you could say hi

<cwebber2-2> hello Natalie!

Natalie: I guess I should give a summary next week, or would you like to have an idea today of how we found it?

Arnaud: I'd liek ot hear it, since we have time if you don't mind
... Maybe not getting into too much technical details that absent people wouldn't want to miss, but more of a higher level would be interesting

Natalie: It was fairly straightforward to use, as just a format for the data. There are a few little things which I'm not sure why they're there and would have changed
... like having an @ on type, it's extra work in clojure and I'm not sure what tha'ts for
... Whether or not it's recommended to use the types in AS2 core, or if it's okay to make up your own, and what's intended, we weren't really sure
... Another thing that we were originally using the words from the types themselves were being displayed to the user, but actually we needed to translate everything
... Based on what the type was, rather than actually using the words themselves
... In general through it was fairly striaghtforward. The examples were good
... Someone has been working on having examples that people can test against, test parsers with, that's great. We've got tests already, but that's brilliant, and having more examples of what it should look like
... I don't have much feedback other than little things
... Another thing would be activities having incrementing ids
... For ordering. Sometimes timestamps are thes ame

<rhiaro> Isn't that an implementation detail?

Natalie: That's pretty much all the things
... Is everyone here familiar/using/involved?

Arnaud: Sounds positive overall, but one question is
... maybe you should mention generally speaking what this project is about
... I'm not sure everybody is aware. What do you actually do with AS2?

Natalie: at Thoughtworks we're building three tools as part of the dcent project. One of those tools, Mooncake, is a notifications tool
... Mooncake consumes AS2
... One of our other tools, Objective8, publishes activities
... an example would be if someone creates a new objective or asks a question
... these are possible actions in Objective8
... That gets published through an API which Mooncake consumes
... Sort of like a twitter feed, creates a new notification for any new objective or question
... The idea of the notificaiton tool is to give users feedback on any activiyt or event having in their dcent ecosystem
... Anything that's publishing activity streams that's connected to their instance of Mooncake

Arnaud: Thank you.

<Zakim> azaroth, you wanted to +1 rhiaro re alias

azaroth: I just wanted to ask a quick question
... Mostly to +1 what Amy noted about @ in irc
... In the most recent version id and type are aliased to @id and @type
... Whether the documentation could be clearer about that that the intend is to use id and type and avoid the @s which would solve the problem

Natalie: Ah thank you. Looking at that now

sandro: does a json-ld processor naturally favour those aliases?

<rhiaro> does the context take care of that?

sandro: in generation, not parsing

azaroth: I believe yes
... The playground does
... To be tested, but..

<azaroth> +1 to Sandro

sandro: so when we have a test suite it should be clear enough that the @ signs don't belong there because you'll get an error

Natalie: why are they there in the examplse?

<cwebber2-2> wait

<cwebber2-2> are @id and id an error in AS2?

sandro: they're in json-ld

<cwebber2-2> or is it just supported to have the alias

<cwebber2-2> sandro, ^^^

sandro: But we changed them to not need it in AS2

<azaroth> cwebber2-2: Good question

Arnaud: do you actually use json-ld, or you read this as json?

Natalie: json

sandro: but you do use extensions right?
... You have your own vocabulary for some of the things going on?

Natalie: What do you mean? I don't think we've added anything that's not in here

sandro: I had the impression from something I heard that when Objective8 is sending it's feed it's using a vocabulary that includes things relative to what Objective8 does, rather than just general AS2 things. Maybe I misunderstood

Natalie: No, just using AS2

Arnaud: This is interesting, touches a point i wanted to make
... I wanted to point out that AS2 is meant to be extensible so you could add your own elements to it, to your stream, so you can carry extra information that is not part of the standard
... Iw as going to ask if you use any extensions, so you just answered that, but I don't know if you're aware that you could

Natalie: surely the idea about the standards that if this is the data I'm publishing I'm assuming that any consumer will only accept the AS2 standard keys
... So if I add my own that data will disappear it won't be used

Arnaud: You're right, any time you use some extension you won't achieve the same level of interoperability that you have otherwise. But there is a sweet spot we're trying to find where we're not trying to cover the world, and at the same time tryign to provide a core set of functionality taht will address most of your use cases, but leave the door open for those who need to go beyond

Natalie: I think a lot of what we were doing was around the usage of AS2 and trying out... I guess if you've got an endpoint that is being consumed by many things you might want that extensibility, but we were only being consumed by Mooncake which had the data in this format

<Zakim> azaroth, you wanted to +1 cwebber's question

Natalie: Making it clear which is the data being consumed by AS2 consumers, and what is being consumed by other things

azaroth: Mooncake is a really awesome name btw

<KevinMarks> it may not disappear - depends on the client. One workaround is to put mf2 in your content

<KevinMarks> hm

azaroth: cwebber2 had a good question
... that @id would be an error?

<KevinMarks> https://www.w3.org/wiki/Activity_Streams links to http://jasnell.github.io/w3c-socialwg-activitystreams/activitystreams2.html whcih 404s

sandro: if both are acceptable, every consumer has to understand both, so natalie's problem is not solved
... Her code has to check for both
... So I was assuming we'd make it an error bu tI don't know
... Something someone should find in the spec, and make a test case

Arnaud: is natalie on irc?
... Not that you have to be
... but peopel make comments and side questions there
... so if you hear people saying somebody else said this or that, they read it on irc
... I realise if you're not on there it might sound a bit weird
... Any other questions for natalie?

azaroth: Is someoen going to take an action to check that? Especially as the examples all say @ and that's an error, that's a CR blocker

Arnaud: was that volunteering?

azaroth: .......sure, I'll do it.

<sandro> no good deed goes unpunished :-)

Arnaud: Thank you

<azaroth> :)

Arnaud: Any other comments or questions?

sandro: I know a little bit of this, but Natalie do you have peopel using these tools now, or expectations about whether they'll be publich or who is using them going forward

Natalie: we don't have users, we do have a live demo
... I can send the links to sandro
... People go and play with that, it's a sandbox
... at the moment we've got people looking at the other tools
... Mooncake is more useful once you've a few things in the dcent ecosystem, at the moment we're only trying to get people to use one thing
... Less useful if it's just one thing
... That being said, lots of people are looking at these tools so we'll see if anyone is interested

Arnaud: If you can send a link I'm sure people would like to have a look

Natalie: Mooncake Demo: http://mooncake.dcentproject.eu/
... Mooncake Code: https://github.com/d-cent/mooncake
... Objective8 Demo: https://objective8.dcentproject.eu/
... Objective8 Code: https://github.com/d-cent/objective8

sandro: How long will you be working on this? I understand dcent is winding down in a little bit

Natalie: At the moment (things change) the dcent project ends at the end of May, including THoughtworks involvement
... Whether that will continue because there's more funding or we want to keep it going I'm not sure, but at the moment our end date is may
... These tools are open source though

<ben_thatmustbeme> https://github.com/ThoughtWorksInc/mooncake

Arnaud: We're tryign to get to CR for AS2, which is when we as a working gropu declare that we have a spec that addresses our use cases and doesn't have any known technical issues, and that's when we're calling the world on implementation feedback
... So hopefully we'll be there before the end of may and then you can provide some formal feedback
... THat would be very helpful
... Let's move on
... I don't know what else we can talk about given the lack of editors
... WAiting for Evan to get back to us on different pieces
... Invited everyone else to add to the agenda, haven't seen any edits. Anything anyone wants to bring up?

KevinMarks: Issues were already on the agenda, and I posted a couple of issues based on external comments, mostly reposting exteral comments
... on github, just to note I did that, to bring in some outside implementors who had questions and issues

Arnaud: do you think the spec needs to be modified?

KevinMarks: They're reactions to attempting to implement, so worth reading, but need clarifying to make actions to amend the spec

Arnaud: Thank you

sandro: Other possible agenda items, I wonder... we're four or five weeks from the f2f
... Hopefully everybody has answered on the wiki whether they're coming
... I assume the chairs will start to think about the agend,a but if people have goals or milestones they'd like to push for and have group support to push for, having some progress on each of our documents around the face to face, before or after, is probably important
... So having the editors think about what's necessary to move their document forward

Arnaud: +1
... there are quite a few document sin draft stages, I don't seem to see much activity, so I don't know what's going on
... You can expect as the f2f gets closer, you will be asked what's going on

<ben_thatmustbeme> for my doc its that i'm super busy

Arnaud: Always a question of whether people are busy which I understand, or if there's something that's blocking progress that we should tackle as a WG
... Unless there's anything else we can close early

sandro: I forwarded Natalie's links, so if you want to go play with Mooncake or Objective8 during your remaining 35 minutes you can do that

Arnaud: Let's close on this, thank you all for joining

Summary of Action Items

Summary of Resolutions

  1. Approved Minutes of 2016-01-26
  2. Approved Minutes of 2016-02-09