#jasnelllooking at the first example, I see a couple of things that need to be clarified
#jasnellmost important, the distinction between the Object and the Activity
#jasnelle.g., your statement "There has been a bit of discussion about whether things like audience targeting (to) would be better off attached to the object rather than the activity; I'm inclined to think there's no metadata related to the creating of an object that can't/shouldn't be attached to the object itself, rendering the activity redundant. I keep asking people of examples that contradict this, but so far nobody has given me one."
#jasnellthere is a subtle but important difference that is based on point of view
#jasnellneither point of view is "more correct" than the other, they're just different and equally valid
#jasnellconsider the two sentences "this article was written by rhiaro" vs. "rhiaro posted an article"
#rhiaroyeah, I deliberately merged them for that post to see what it would look/feel like, and to provoke feedback
#jasnelldepending on the article, these can be equivalent statements
#jasnellwhen I use the form {"@type":"Post", "actor": "http://example/rhiaro", "object": {"@type": "Article", ...} }, I'm talking about two separate things: The Article, and the Activity of Posting the Article
#jasnelleach can have their own sets of properties
#jasnellfor instance, you may have written the article a week ago but only posted it today
#elf-pavlikcwebber2, re: "I'm converting the whole activitystreams vocabulary into classes" will instances still suport use of any properties - not defined in AS2 vocab but via extensions?
#jasnellthe article may have it's own intended audience, but the statement "rhiaro posted this article" might be directed at a *different* audience than who the article was originally targeted at
#elf-pavlik"In Linked Data, properties are, just as everything else, identified by IRIs and thus have global scope which implies that they have independent semantics. In contrast, properties in data models as used in common programming languages are class-dependent. Their semantics depend on the class they belong to. In data models classes are typically described by the properties they expose whereas in Linked Data properties define to which classes they belong
#elf-pavlik. If no class is specified, it is assumed that a property may apply to every class."
#tantekelf-pavlik: I keep hearing hydra but not seeing anyone deploy it on their personal site
#cwebber2tantek: I was asking elf-pavlik and jasnell what the official AS2 position is on it ;)
#jasnellrhiaro: the point is, each Activity is it's own distinct object, with it's own sets of properties that are independent of the Object of the activity
#cwebber2I'm messing around with implementing a library for AS2 modification
#cwebber2but I figured I might be missing something in the spec
#jasnellAS2 does take a position on it, a very clear one
#tantekcwebber2 - there's so much handwaving noise in the "social web" space that the only way to actually get things done is to do heavy filtering
#rhiaroright, a repost of an article - whether by the same author or not - is an object of it's own with it's own post date. So it's two objects, rather than an activity and an object. But I the case for wanting a written date and a published/posted date separately makes sense
#jasnellok, so it's a combination of "Article" and the existing "Post" activity type, it conflates two separate concepts: The Thing and the Activity that created the Thing
#rhiaroI understand the difference between activities and objects according to AS2, I'm just not sure they need to be different
#rhiaroI think pushing around objects with all the metadata attached *might* be enough. I'm actively seeking examples where this *isn't* the case, and whilst I obviously haven't covered the entire social web yet, I'm working on it
#tantekrhiaro: the answer is no, no need for two names for things which look, and act the same
#tantekrhiaro: I'm with you I reject this assertion: "you cannot conflate all Activities and Object types" because so far we've figured out how to do so in every case in the indieweb
#rhiaroelf-pavlik: ^ experimental and doesn't do everything I need yet, was just a start
#tantekbasically, shipping various kinds of posts is beating the theoretical arguments for activities abstractions
#rhiarothey're just notes with startTime, endTime, startLocation and endLocation
#tantekthus if *any* post types are a priority for you, you'll find you can implement them more quickly if you start with solving notes for your own site
#aaronpkelf-pavlik: think of "notes" as a minimum viable event and place post
#tantekplus with plain text first design, all post types start as notes
#aaronpkthe note text being the plaintext fallback of any other more complex post
#elf-pavlikaaronpk, plaintext fallback makes sense
#jasnelltantek: excercise... model, "On May 2nd, 2015, Joe and Sally each bought three bottles of 2013 Santa Margherita Pinot Grigio" as a post that captures all the relevant semantic detail
#tantekelf-pavlik: if it's not published in HTML, it's not on the web
#jasnellin other words, I need to know that there are two actors, that they each bought something, what that something was, in what quantity, and on what dates
#aaronpkelf-pavlik: "Error: Failed to lookup view "undefined" in views directory"
#rhiarobtw jasnell I have an activitystream json at rhiaro.co.uk/activitiesjson.php. It's simple and has as little nesting as possible, using only URIs, but I think it should be spec compliant
#aaronpkjust make that return the text of the posts with <br> or something :)
#elf-pavlikaaronpk, i didn't create templates to render HTML yet, not high priority for me!
#jasnelltantek: that's a cop out. imagine that the statement hasn't been posted yet so there's no url created for it yet
#jasnellI'm modeling the statement to post to an API so that the url can be created
#tantekjasnell: not worth considering such theoretical statements given the amount of concrete publishing experience we have
#jasnellbut please, humor me, stretch that imagination just a bit, how would it be modeled *assuming* that it was not theoretical
#tanteknah, that's science 101, real world evidence trumps theories
#tantekreal world examples > hypothetical exercises
#tantekI'm starting to believe that people responding with "in your opinion only" are basically admitting to defeat in a reasoned argument.
#rhiarojasnell: could you post the AS json for that in a gist and I'll see if I can think of a post-y alternative?
#elf-pavliktantek, as for now IndieWeb posting AFAIK can't capture difference between "eat an apple pie" and "bake an apple pie", neither "eat an apple" or "pick an apple from the tree" http://mundraub.org/
#jasnellrhiaro: sure, I can mock it up, give me a few
#tantekelf-pavlik: when you start posting those as notes to your own site with permalinks, let me know so we can study them as real world examples
#rhiaroSo far it sounds like it's probably two posts, one each for joe and sally, that could be combined in the UI but should be modelled separately
#tantekelf-pavlik: anyone can publish theoretical hierarchies of types - not interesting. find some permalinks and then it's worth our time, not til then
#rhiaroI'm going to post offers and requests as POSSEing to Freegle/Freecycle soonish, and I'm pretty sure they're just gonna look like posts
#jasnellthe object types are made up because I don't want to go looking for an existing vocabulary for those types right now
#rhiaroif you wanted to move your twitter chess to a different network, there's no markup to let the new network know what it is beyond strings, if say you discovered a text-based-chess network that could give you stats on how you play if it knows what all your moves are
#cwebber2tantek: I guess the side of games that posts don't feel sufficient for me are those with side effects
#rhiarocan you do partial updates, or do you have to ship the whole object even the properties that aren't changed?
#rhiaro(this is a general question I've been meaning to ask)
#rhiaroie. if you exclude properties that exist in the update, does it remove them or leave them the same?
#jasnellbtw, if you do not need to provide metadata about the activity itself, it's perfectly ok not to
#rhiaroif the activity doesn't even have metadata, why have it at all?
#jasnellthere's nothing in AS2 that requires you to always use the Activity object
#jasnelljust like there's nothing in AS2 that requires you to use all of the other object types
#rhiaroso it *would* be compliant with the current AS2 spec to just publish generic objects?
#jasnellin other words, AS2 allows you to express both statement forms, "Sally posted an article" and "This article was written by Sally" can each be represented equally
#jasnellAS2 is *just* a data format, other than defining syntax validation rules, it does not define conformant behaviors
#hhalpinIn terms of licensing, this is a much larger issue - i.e. we get patents to specs
#hhalpinso that obviously if the specs change/fork, that's an issue re the patent commits (i.e. we'd have to do them again!)
#hhalpinHowever, in terms of forking for the use in examples, etc.
#hhalpinThe W3C is working very hard on loosening that - I think Rigo and Wendy are working with the AC and the Process CG
#jasnellwhen I say, {"@type": "Like", "actor": "http://example.org/rhiaro", "object": "http://example.org/foo", "content": "this is an awesome post"}, the statement itself is an Object, just with specific semantic detail included
#hhalpinsince obviously its useful to have flexibility in using the spec, as long as the patent commits are to a clear text.
#aaronpkhhalpin: we set up the github repos and clarified the difference between the official w3c repos that only the editors can contribute to, and the more free-for-all w3csocial that anyone can contribute to
#aaronpktantek: for new people joining the group, is there a link to that clarification that we can put in this issue and on the home page?
#trackbotaction-60 -- Pavlik elf to Draw Follow vs. Subscribe with account having multiple feeds allowing subscription independently -- due 2015-05-26 -- PENDINGREVIEW
#aaronpk.. elf harry do you have opinions that warrant opening this?
#aaronpkelf-pavlik: to clarify, we don't talk about namespace in jsonld context, it's worth keeping this issue open because we don't have a strategy of how to manage jsonld context
#bblfishok, so I don't know how contexts work in json-ld work
#aaronpktantek: elf it sounds like you think this is enough of an issue that you want it assigned to you
#hhalpin_My two cents is if we do a breaking versiioning change, we just use normal W3C process.
#aaronpkjasnell: what we're talking about is the normative context document. nothing stops implementers from creating their own modified version of it, that extends but does not modify the base terms
#aaronpk.. the fact that our default one only deals with AS2.0 vocab we aren't making any backward compatible changes to the vocab past CR, so we don't need version number there
#aaronpk.. i'd like to request you contact julien directly to try to resolve this
#aaronpkhhalpin_: i've already done this once, bradfitz already said no
#aaronpktantek: because julien is editing this now
#cwebber2> eprodrom: In previous telecons we've gone over raised issues, but that's been controversial. What I'd like to do is do that at the end of the agenda if we have time
#aaronpkhhalpin_: it doesn't matter if julien is editing, becasue it refers to a previous document that we don't have any licensing on
#aaronpkhhalpin_: dual licensing has been discussion, it's beyond this WG to do anything, the answer right now is no, but the answer in the future is likely to be yes
#aaronpk.. there's nothing this group can do, it's a w3c issue
#Zakimsees elf-pavlik, bblfish on the speaker queue
#wilkierhiaro: just experimental. i don't really know what I could add other than it just feels like when i put micropub stuff into AS it was just objects in AS and that felt simpler
#wilkieeprodrom: I guess I'm finding the conversation pretty interesting. I think there is definitely an intense cluster around object CRUD lifecycle (create, read, update, delete) a particular object (note etc)
#wilkieeprodrom: I hate to be the stickler, but I think that for developers outside of the call, if we publish something that doesn't say this is how we post a note, etc, but rather this is how you do a thing without tying to use-cases, I would be worried it wouldn't be actionable outside of us.
#wilkieeprodrom: we need some decisions around vocabulary in there
#hhalpin_Agreed, we'll need to tie it all the use-cases and specify examples with vocabularies *before* FPWD publication
#eprodromLet's just makes sure we don't publish the "Social API that sends a thing to a place to do some stuff"
#wilkietantek: I agree if people saw an abstract draft that it would not look good
#hhalpin_If the big vocabulary difference is microformats vs. AS 2.0 Vocabulary, we can move to that discussion next - as I think elf was interested in RDF versions of microformats.
#wilkiebblfish: I wonder if the vocabulary could be put different. I think the issue is what side-effects exist for publishing some kind of thing. if you post a picture, what are you liable to deal with to do that.
#rhiarohhalpin_: also yeah we should do better at documenting our (me aaron jessica) discussions about form-encoding and json, we've thought a bit about this
#wilkiebblfish: that's the one thing I feel isn't addressed here, but perhaps I just looked too quickly
#hhalpin_I would be happy to see form encoding, as I think developers want it, although REST folks might be unhappy
#wilkieelf-pavlik: I collected existing efforts and I posted awhile ago the invitation for people could subscribe to each other using these different formats
#Zakimsees elf-pavlik, hhalpin_ on the speaker queue
#elf-pavlikInvitation (challenge) addressed at two of Social WG chairs - Tantek & Evan proposing Proving usability of ActivityStreams2.0 and Microformats together by example
#wilkieeprodrom: not interested in participating in this challenge. I find it condescending and rude. not interested in completing arbitrary challenges. please do not bring it up again. thank you.
#Zakimsees elf-pavlik, hhalpin_ on the speaker queue
#eprodromI'm mostly frustrated that it's personalized
#wilkietantek: I will point out for the record that previously eprodrom and I have gotten interoperability. there has been no problem in doing so. I'm also not immediately interested in this type of challenge.
#wilkieelf-pavlik: apologies to eprodrom. I thought it would be a pragmatic way to address issues of interop. I will close the github issue and not bring it up again.
#wilkieelf-pavlik: my question is that if we had clear prefix URIs we could use them in both. anyone who wants to use them can just use them.
#wilkiehhalpin_: I think that's a reasonable way forward. there seems to be a lot of overlap. we have to make a choice. a good way is to choose one and make a union and where there is an overlap use microformats terms instead of making new ones.
#wilkiehhalpin_: I don't think it really needs an implementation challenge, we should just make a decision on how we will handle these two vocabularies
#cwebber2hhalpin_: I'm all for implementation union etc
#hhalpin_Another way to go forward would just be to merge all stable microformats into AS2.0, and then if there's overlap go with microformats term as that already has wide deployment.
#wilkietantek: is that a good enough incremental progress for us to move on?
#cwebber2hhalpin_: btw pump does already have deployment, so
#hhalpin_Yes, but AS2.0 is still I think a bit more unstable
#Zakimben_thatmustbeme, listening for 10 seconds I heard sound from the following: jasnell (15%), Marilyn (15%)
#wilkieeprodrom: I think the idea with as:Profile is that it would be an object (not an actor) that you would be able to do things like update or delete.
#wilkieeprodrom: the idea was that instead of having a person object update that same person object to represent profile update, we would have person update profile objecct
#wilkieeprodrom: yes, elf-pavlik, those are similar
#wilkieeprodrom: we also had tangled in there the concept that you might have a user account with multiple profiles
#wilkieeprodrom: or rather a person with multiple profiles so I can have a work profile, friends profile, or maybe a political profile
#tantek"I-JSON (short for "Internet JSON") is a restricted profile of JSON designed to maximize interoperability and increase confidence that software can process it successfully with predictable results."
#wilkiejasnell: it is possible that JSON-LD implementations to not ~output~ I-JSON compatible documents
#hhalpin_We'd have to do some double-checking in the test-suite that we were I-JSON compliant - it's a small issue, but Erik Wilde supported it as a best practice and I tend to do agree.
#wilkiejasnell: AS implementations should try to make I-JSON compatible documents. if consensus suggests we should make an explicit note about it, I'm happy to do that.
#ZakimAs of this point the attendees have been Ann, jasnell, +1.314.705.aaaa, Arnaud, Marilyn, aaronpk, elf-pavlik, +1.401.305.aabb, ben_thatmustbeme, Sandro, rhiaro, AdamB, tantek,
#ZakimAs of this point the attendees have been Ann, jasnell, +1.314.705.aaaa, Arnaud, Marilyn, aaronpk, elf-pavlik, +1.401.305.aabb, ben_thatmustbeme, Sandro, rhiaro, AdamB, tantek,
#jasnellelf-pavlik: yeah, I was thinking of externalizing the examples in a separate JSON file, assigning each a UUID, then using a simple js to pull those into the text, then when I actually cut the WD, I would pull the generated html using the respec tools. It should be fairly straightforward. What we would need to do tho is replace the inline examples with a simple placeholder referencing the appropriate UUID.
#trackbotaction-57 -- Pavlik elf to extract all AS2.0 examples in specs into separate files + expose them to JSON-LD Playground -- due 2015-04-14 -- CLOSED
#trackbotaction-50 -- Evan Prodromou to Extract the examples from the main documents, pick properties and put together a test for those examples -- due 2015-03-24 -- OPEN
#jasnellobviously there are use cases there that go beyond our core user stories, but it should be possible for implementers to make use of AS2 for these kinds of scenarios
#tantekjust uses "uid" instead of UUID because of that.
#jasnellcwebber2: looking at the user stories list, there's only one that mentions "confirm" but it's not clear in that context if it's any different than "Accept"
#jasnellcwebber2: so given that, let's just drop it
#melvsteri noticed this one is 10 years and 10 days old today :
#melvster Your global identifier throughout the web isn't "happygirl234324" or an email address, or "bradfitz@identityserver.com", but your FOAF URL. So you also choose how much info you do or do not want to share in there.