#social 2015-06-30

2015-06-30 UTC
bblfish and jasnell joined the channel
#
Loqi Riannell made 1 edit to [[Socialwg/invited-experts]] https://www.w3.org/wiki/index.php?diff=84897&oldid=84818
jasnell, bblfish, tilgovi, jaywink, melvster and shepazu joined the channel
#
melvster rhiaro: you may be interested in this post https://lists.w3.org/Archives/Public/semantic-web/2015Jun/0125.html
#
melvster it talks about predicates without a namespace in relation to linked data
#
melvster my idea was when you see an unnamespaces key like "predicate" to change it into linked data using "urn:predicate"
#
melvster s/unnamespaces/unnamspaced
#
melvster it's even easier in json ld you can just have "@vocab" : "urn:"
#
melvster and everything is automagically converted
#
rhiaro melvster: wouldn't the document URI make sense as the namespace where none is set, as that document is the context in which the predicate is being used? Someone could use the same predicate in different documents to mean differnet things, so merging them with urn: wouldn't be a good idea?
csarven joined the channel
#
elf-pavlik hey! does your conversation relate to https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134 ?
tantek joined the channel
#
melvster rhiaro: yes I used to do that a lot, only issue is that the names are then not consistent over multiple docs
#
melvster perhaps the document uri already is the default
#
rhiaro it becomes a toss up between collisions and inconsistency
#
melvster true!
#
rhiaro one might cause wrong inferences, one not enough inferences
#
rhiaro I think wrong inferences is worse
#
melvster elf-pavlik: what the: "@vocab": "_:" hack!
#
elf-pavlik https://github.com/jasnell/w3c-socialwg-activitystreams/blob/master/activitystreams2-context.jsonld#L3
#
melvster elf-pavlik: could you summarize what it does?
#
elf-pavlik previous conversations around that also in: https://github.com/jasnell/w3c-socialwg-activitystreams/issues/36
#
elf-pavlik also check out http://tinyurl.com/nas73bc
#
elf-pavlik if term e.g. *foo* doesn't have mapping in JSON-LD context, AS 2.0 will use assign it _:foo as identifier
eprodrom joined the channel
#
melvster ah
#
melvster clever
#
elf-pavlik i explained issue with some "wrong interpretation" in this comment https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134#issuecomment-108255400
#
melvster rhiaro: I think you need to add a # to make the predicate relative to the document, right?
#
melvster to my mind a variable is a name
#
melvster if you dont have any other information and you want to convert it to a URI, urn: makes sense to me, as it's designed for names
#
elf-pavlik csarven, hey! glad to see you joining this WG :)
#
melvster csarven++
#
Loqi csarven has 8 karma
#
csarven :)
#
melvster if we consider a raw variable or raw json key to be 'just a name' ... tha tis
#
melvster just seemed sort of logical to me
#
melvster i didnt even realize AS2 had a default vocab, elf-pavlik, thanks for pointing that out!
#
elf-pavlik melvster, IMO normative JSON-LD context for AS2.0 doesn't get enough visibility and as much review as core spec and vocab spec
tilgovi joined the channel
#
melvster well _: is pretty smart I have to say
#
melvster i would never have thought of that
#
melvster csarven: maybe that's the solution ^^ _:
bblfish joined the channel
#
melvster elf-pavlik: is it considered a 'hack' for any reason?
#
csarven bnode?
#
elf-pavlik melvster, have you read my comment with ACME and Baz Inc?
#
csarven doesn't know the context of the discussion.
#
aaronpk bnode--
#
Loqi bnode has -1 karma
#
melvster elf-pavlik: so you want to use : "@vocab": "http://example.org/ns/" ?
#
elf-pavlik csarven, https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134
#
csarven Poor bnodes. They always take the first beating.
#
melvster aaronpk: you dont like bnodes?
#
aaronpk those are the things that don't have URLs right?
#
melvster yes
#
elf-pavlik melvster, i propose that if someone wants to use custom terms, person should just create namespace for them. even if it will just give HTP 404 to get started
#
aaronpk then yes
#
aaronpk urls++
#
Loqi too much karma!
#
melvster aaronpk: everything *starts* as a bnode
#
elf-pavlik this way you don't end up with merging two AS2.0 documents and having same identifier (blank node identifier _:foo) for two different terms
#
elf-pavlik issue 134 caputes it with and example merging documents published by ACME and Baz Inc.
#
trackbot doesn't understand that ISSUE command.
#
elf-pavlik i think we may need a NOTE in the end, which explains how to use AS2.0 as RDF
#
csarven aaronpk Well, in mf, IIRC, everything is a bnode (if mapped to RDF). mf's "URLs" just happen to be a property of that thing. They [URL] are not the identifier for that thing.
#
melvster aaronpk: we should name everything, but in the beginning nothing is named
#
csarven Perhaps that's just schemantics.
#
aaronpk i was pretty sure the URL was the identifier for the thing
#
csarven In any case, nowadays it is usually agreed that bnodes are a last resort solution.
#
melvster url is a locator
#
aaronpk schemantic! lol
#
melvster uri is an identifer
#
melvster i try to avoid *producing* bnodes
#
melvster but it's hard to avoid *consuming* them
#
csarven I gave up on bnodes completely. Totally fine to put an HTTP fragment out there which may or may not exist. Deal with it later.
#
rhiaro dealwithitlater++
#
rhiaro :)
#
elf-pavlik for example *p3k-drink* ('pseudonamespaced' term extending microformats) doesn't have URL http://pin13.net/mf2/?url=https%3A%2F%2Faaronparecki.com%2Fmetrics%2F2015%2F06%2F29%2F205422%2F
#
melvster there's actually bnodes in the AS2 spec
#
aaronpk elf-pavlik: that post has a URL
#
elf-pavlik aaronpk, post yes but not vocab term: p3k-drink
#
aaronpk meh
#
aaronpk it's free-form text
#
rhiaro aaronpk: that's what the conversation is about - property names
#
aaronpk not a vocabulary
#
rhiaro in RDF they all have their own namespaced identifiers
#
rhiaro so that when two people use the same property name, you know they mean the same thing
#
csarven rhiaro in theory :)
#
melvster rhiaro: but they dont have to, if you think of properties as names, and it makes sense to have urn:
#
csarven People misuse properties all the time.
shepazu joined the channel
#
melvster csarven++
#
Loqi csarven has 9 karma
#
elf-pavlik at least properties have *globaly unique* names, even if misused
#
rhiaro csarven: I wrote a paper about people misusing property names all the time :)
#
melvster rhiaro: i think it's actually impossible *not* to misuse properties
#
ben_thatmustbeme tries to follow conversation this late in
#
melvster 16 of the 24 AS2 examples contain bnodes
#
csarven There must be a correlation between data quality (specifically proper use of vocabs/onts) and vocab/ont quality. <-- Who wants to do a PhD?
#
ben_thatmustbeme difference is mis-use of standardized properties vs mis-use of extension properties (if i follow that correctly).
#
melvster elf-pavlik: what does the _: prefix mean?
#
elf-pavlik blank node
#
elf-pavlik http://www.w3.org/TR/json-ld/#identifying-blank-nodes
#
elf-pavlik "The blank node identifier is scoped to the document in which it is used." IMO it can come tricky when merging two AS2.0 documents which happen to use same custom property which gets mapped to same _:foo while having intended different meaning
#
elf-pavlik does anyone know what time earliest i can dial in to check Webex connection?
#
ben_thatmustbeme no clue
#
rhiaro elf-pavlik, how are you planning to connect?
#
elf-pavlik i will use friend's iPad with Webex software
#
rhiaro oh, cool
#
rhiaro I installed the client on my phone, but no audio last week, for no apparent reason
#
elf-pavlik hmmm...
#
tsyesika is there an echo test for WebEx?
#
elf-pavlik i'll try to call now to see what happens
#
elf-pavlik i get 'meeting not started yet'
#
elf-pavlik aaronpk, ben_thatmustbeme do you have any examples of *access controll* on HTML pages (aka read API) in IndieWeb community?
#
elf-pavlik i stay again with fkooman in Berlin and we'll play with it more in next days :)
#
aaronpk elf-pavlik: http://indiewebcamp.com/private_posts
#
melvster elf-pavlik: yes I think you're right, merging two documents then does the wrong thing
#
melvster blank nodes are not supposed to collide
#
melvster but urn: can
#
elf-pavlik aaronpk, as far as i remember nothing that would work with a 'reader' ?
#
elf-pavlik http://indiewebcamp.com/private_posts#FAQ
#
melvster elf-pavlik: do you know if it is possible to override the @vocab attribute with another @vocab?
#
aaronpk afaik nobody has built that yet but it's a small step to get your reader to be able to authenticate as you when fetching a feed
#
melvster elf-pavlik: and are bnodes being used as predicates?
#
elf-pavlik "The serialized JSON form of an Activity Streams 2.0 document must be consistent with what would be produced by the standard JSON-LD 1.0 Processing Algorithms and API [JSON-LD-API] Compaction Algorithm using, at least, the normative JSON-LD @context definition provided here. Implementations may augment the provided @context with additional @context definitions but must not override or change the normative context. "
#
elf-pavlik http://jasnell.github.io/w3c-socialwg-activitystreams/activitystreams2.html#syntaxconventions
#
melvster < dlongley> JSON-LD supports blank nodes as properties, standard RDF does not, but generalized RDF does.
#
elf-pavlik yeah, i remember very well this article http://manu.sporny.org/2013/rdf-identifiers/
#
melvster elf-pavlik: the author of that article hates the semantic web
#
elf-pavlik melvster, In Your Opinion :)
#
sandro Action: sandro to move forward modified WG Charter, just changing the document license
#
trackbot is creating a new ACTION.
#
trackbot Created ACTION-70 - Move forward modified wg charter, just changing the document license [on Sandro Hawke - due 2015-07-07].
#
elf-pavlik sandro++
#
Loqi sandro has 18 karma
#
melvster elf-pavlik: nope he wrote an article "Why I Hate the Semantic Web"
#
melvster http://manu.sporny.org/2014/json-ld-origins-2/
#
elf-pavlik needs to read it!
#
melvster haters gonna hate :)
#
csarven I presume we put our IRC nick in MIT webex thingy?
#
wseltzer csarven, the important thing is to present+ in IRC
#
csarven Well, I'm having a 'java issue' when I try to join at https://mit.webex.com/mit/j.php?MTID=m60d8f99fc911f2b6fdbb0650665f22b4
#
csarven Using Firefox/Ubuntu.
#
wseltzer csarven, you can just dial in by phone without the webex client
#
csarven Cool, will do that. Thanks.
#
csarven Google Talk/Voice was working earlier.
#
aaronpk where is the info for getting the osx client?
#
aaronpk didn't see it on the meeting agenda or our home page
Arnaud joined the channel
#
elf-pavlik aaronpk, https://www.w3.org/2006/tools/wiki/InstallingWebEx
#
elf-pavlik i just connected, could someone please already join so i can test voice?
#
aaronpk still doesn't understand why a brand new page on w3.org would have "2006" in the url :(
#
Loqi it'll be ok
#
wseltzer datespace, aargh
#
csarven aaronpk Why do you assume that it is a year? ;)
#
aaronpk heh
#
csarven Loqi++
#
Loqi Loqi has 354 karma
#
csarven Everything is going to be aaaalright.
#
aaronpk cause it's a recent year and is the first part of the path which is a pretty common pattern for URLs?
#
aaronpk huh webex says "this meeting has been cancelled"
#
elf-pavlik did anyone join webex call aready ?
#
csarven something URLs something opaque something
#
elf-pavlik 642 112 960 ?
#
melvster is that a year?
#
aaronpk that page is not helpful, it tells me to install java
#
elf-pavlik http://www.w3.org/TR/webarch/#uri-opacity
#
aaronpk wild guessing... goes to mit.webex.com
#
elf-pavlik aaronpk, https://itunes.apple.com/us/app/cisco-webex-meetings/id298844386
#
aaronpk worked this time
#
aaronpk turned it off and back on again
#
aaronpk webex plugin wants to take over chrome :P
#
cwebber2 hey elf-pavlik !
#
elf-pavlik hey cwebber2 :)
#
cwebber2 nice to have you back, I think you were away the last couple?
#
cwebber2 or my memory was away!
#
cwebber2 one of them :)
#
elf-pavlik who joined as Call-in-User_2 ?
#
csarven I don't know my call-in-user #
#
aaronpk i miss zakim
#
elf-pavlik csarven, i guess you than :)
#
csarven present+
#
cwebber2 I just joined
#
elf-pavlik present+ elf-pavlik
#
ben_thatmustbeme ben_thatmustbeme+
#
ben_thatmustbeme or wait, was it
#
ben_thatmustbeme +ben_thatmustbeme
#
tantek present+ tantek
#
aaronpk don't we need to wait for trackbot to start the meeting before doing present+?
#
csarven present+ csarven
#
sandro present+ sandro
#
eprodrom present+ eprodrom
#
cwebber2 present+ cwebber2
#
Arnaud present+ Arnaud
#
wseltzer rrsagent, pointer?
#
sandro trackbot, start meeting
#
trackbot is preparing a teleconference.
RRSAgent joined the channel
#
RRSAgent logging to http://www.w3.org/2015/06/30-social-irc
#
trackbot RRSAgent, make logs public
#
RRSAgent I have made the request, trackbot
Zakim joined the channel
#
trackbot Zakim, this will be SOCL
#
Zakim ok, trackbot; I see T&S_SOCWG()1:00PM scheduled to start now
#
trackbot Meeting: Social Web Working Group Teleconference
#
trackbot Date: 30 June 2015
#
cwebber2 presentbot? :)
#
aaronpk present+ aaronpk
#
sandro present+ sandro
#
jasnell parent+ jasnell
#
elf-pavlik https://www.w3.org/2006/tools/wiki/WebExBestPractices#Attendance
#
jasnell present+ jasnell
#
cwebber2 present+ cwebber2
#
eprodrom ?
#
Arnaud present+ Arnaud
#
ben_thatmustbeme hears the bells
#
eprodrom They're kind of charming
#
aaronpk ah it's RRSAgent
#
Arnaud yes, you have to do it after rrsagent joined
#
wseltzer present+ csarven, elf-pavlik, ben_thatmustbeme, eprodrom, tantek
#
csarven present+ csarven
#
wseltzer thinks she got the people who came in earlier
#
tantek zakim?
#
tantek zakim, who is here?
#
Zakim T&S_SOCWG()1:00PM has not yet started, tantek
#
Zakim On IRC I see RRSAgent, Arnaud, bblfish, eprodrom, tantek, csarven, jasnell, melvster, jaywink, KevinMarks, elf-pavlik, wilkie, Loqi, cwebber2, slvrbckt, sandro, trackbot, kylewm,
#
Zakim ... raucao, ben_thatmustbeme, tsyesika, aaronpk, dwhly, JakeHart, bigbluehat, ElijahLynn, tessierashpool_, mattl, oshepherd, tommorris_, bret, rhiaro, wseltzer
#
tantek trackbot, start meeting
#
trackbot is preparing a teleconference.
#
trackbot RRSAgent, make logs public
#
RRSAgent I have made the request, trackbot
#
trackbot Zakim, this will be SOCL
#
Zakim ok, trackbot; I see T&S_SOCWG()1:00PM scheduled to start 3 minutes ago
#
trackbot Meeting: Social Web Working Group Teleconference
#
trackbot Date: 30 June 2015
#
bigbluehat present+ bigbluehat
#
tantek Zakim, who is here?
#
Zakim T&S_SOCWG()1:00PM has not yet started, tantek
#
wseltzer present+ wseltzer
#
Zakim On IRC I see RRSAgent, Arnaud, bblfish, eprodrom, tantek, csarven, jasnell, melvster, jaywink, KevinMarks, elf-pavlik, wilkie, Loqi, cwebber2, slvrbckt, sandro, trackbot, kylewm,
#
Zakim ... raucao, ben_thatmustbeme, tsyesika, aaronpk, dwhly, JakeHart, bigbluehat, ElijahLynn, tessierashpool_, mattl, oshepherd, tommorris_, bret, rhiaro, wseltzer
#
wseltzer alas, poor zakim
#
Arnaud zakim, won't do much more than manage the queue I think
#
Zakim I don't understand you, Arnaud
#
aaronpk poor Zakim
#
elf-pavlik Benjamin Young - i don't recognize your IRC handle...
#
ben_thatmustbeme what is left that we use from Zakim?
#
aaronpk just q+
#
tantek q?
#
Zakim sees no one on the speaker queue
#
ben_thatmustbeme elf-pavlik Benjamin Young is bigbluehat
#
tantek https://www.w3.org/wiki/Socialwg/2015-06-30
#
elf-pavlik thanks ben_thatmustbeme
#
elf-pavlik i can scribe
#
tantek scribe: elf-pavlik
#
ben_thatmustbeme scribeNick: elf-pavlik
#
elf-pavlik agenda: https://www.w3.org/wiki/Socialwg/2015-06-30
#
tantek https://www.w3.org/wiki/Socialwg/2015-06-23-minutes
#
elf-pavlik topic: Approval of Minutes of 16 June 2015
#
bigbluehat tnx ben_thatmustbeme
#
cwebber2 rhiaro++
#
Loqi rhiaro has 109 karma
#
tantek rhiaro++ for helping convert the minutes to the wiki
#
cwebber2 (for putting up minutes)
#
Loqi rhiaro has 110 karma
#
eprodrom +1
#
elf-pavlik s/16 June 2015/23 June 2015/
#
ben_thatmustbeme +1
#
tantek topic: https://www.w3.org/wiki/Socialwg/2015-06-30#Charter_License_Update
#
elf-pavlik we decided to use new W3C license for all the work from this group
#
elf-pavlik tantek: sandro have accepted this action item and will take it through W3C process
#
cwebber2 sandro++
#
Loqi sandro has 19 karma
#
elf-pavlik ... sandro do you have any rough estimates how long it will take?
#
elf-pavlik sandro: it should take 2-3 to take it to AC
#
elf-pavlik ... feel free to remind me next week
#
elf-pavlik tantek: any questions on this point?
#
tantek Topic: https://www.w3.org/wiki/Socialwg/2015-06-30#Activity_Streams_2
#
elf-pavlik tantek: we have two drafts prepared for publication
#
jasnell q+
#
Zakim sees jasnell on the speaker queue
#
elf-pavlik ... thank you jasnell for preparing them! also ben_thatmustbeme thanks for PR with microformats examples
#
tantek ack jasnell
#
Zakim sees no one on the speaker queue
#
ben_thatmustbeme i have more MF2 updates to do
#
elf-pavlik jasnell: eprodrom posted number of PRs with grammar and spelling corrections which will get included
#
ben_thatmustbeme but won't be ready for a little while
#
KevinMarks present+
#
Loqi Tantekelik made 1 edit to [[Socialwg/2015-06-30]] https://www.w3.org/wiki/index.php?diff=84899&oldid=84870
#
ben_thatmustbeme present+
#
elf-pavlik tantek: you have purely editorial updates which you have included and which people might not read since they arrived this morning
#
elf-pavlik jasnell: correct
#
eprodrom q+
#
Zakim sees eprodrom on the speaker queue
#
elf-pavlik jasnell: ben_thatmustbeme do you have batch of updates to examples ready?
#
elf-pavlik ben_thatmustbeme: not for today
#
elf-pavlik tantek: what precent have you gotten through ben_thatmustbeme?
#
elf-pavlik ben_thatmustbeme: i belive over half, most from core spec but vocab spec has a lot of examples as well
#
tantek ack eprodrom
#
Zakim sees no one on the speaker queue
#
jasnell my proposal then would be to publish as is currently and push the next round of mf example updates to the next WD
#
elf-pavlik eprodrom, i've done a number of editorial PRs and have more through out rest of the day, i'll leave it to jasnell to decide which he wants to include before publishing new working drafts
#
elf-pavlik ... i don't think anything that should hold publishing new WDs
#
jasnell eprodrom++
#
Loqi eprodrom has 17 karma
#
tantek PROPOSAL: Publish AS2 and AS2 vocab drafts as is currently with a few more purely editorial edits from eprodrom today, and push next round of mf examples updates to the next WD.
#
elf-pavlik +1
#
jasnell +1
#
tantek +1
#
cwebber2 +1
#
aaronpk +1
#
sandro +1
#
ben_thatmustbeme +1
#
eprodrom +1
#
tsyesika +1
#
tantek RESOLVED: Publish AS2 and AS2 vocab drafts as is currently with a few more purely editorial edits from eprodrom today, and push next round of mf examples updates to the next WD. jasnell to prepare documents and submit to pub process by tomorrow.
#
Arnaud q+
#
Zakim sees Arnaud on the speaker queue
#
elf-pavlik jasnell++
#
Loqi jasnell has 22 karma
#
tantek ack Arnaud
#
Zakim sees no one on the speaker queue
#
eprodrom q+
#
Zakim sees eprodrom on the speaker queue
#
elf-pavlik Arnaud: i have question with regard to next step, jasnell do you have a sense what it will take to go from current drafts to Candidate Recomendation (CR)
#
tantek q?
#
Zakim sees eprodrom on the speaker queue
#
elf-pavlik jasnell: we need aditinal review on extended vocabulary e.g. possibly adding *Organization* back as Actor type also an *insturment* proeprty
#
tantek q?
#
Zakim sees eprodrom on the speaker queue
#
elf-pavlik ... examples review as ben_thatmustbeme is doing
#
elf-pavlik ... editorial review as eprodrom is doing
#
tantek ack eprodrom
#
Zakim sees no one on the speaker queue
#
elf-pavlik Arnaud: it answers my question
#
elf-pavlik eprodrom: i see it very likely that Social API which we will come up with will use in some ways AS2.0, i'm not sure if we should go to CR unless we have better idea how Social API will look like
#
tantek +1 to eprodrom's point.
#
tantek q?
#
Zakim sees no one on the speaker queue
#
elf-pavlik ... we may see some aditional needs as we get further with specs of Social API
#
rhiaro +1 to eprodrom
#
rhiaro is lurking
#
eprodrom Agree
#
elf-pavlik jasnell: the impact Social API may have the most would be the vocabulary
#
elf-pavlik ... we could push core to CR and hold on with vocab for little longer
#
Arnaud q+
#
Zakim sees Arnaud on the speaker queue
#
tantek ack Arnaud
#
Zakim sees no one on the speaker queue
#
elf-pavlik ... i would rather not wait for to long
#
elf-pavlik Arnaud: from process point of view, we have 3 requirement - 1. resoved all the issue 2. test suite 3. broad review
#
sandro q+
#
Zakim sees sandro on the speaker queue
#
elf-pavlik ... we should aim at closing all the issues, working on test suite and get broad review while we work on Social API
#
elf-pavlik tantek: Arnaud is correct about the CR requirements from process perspective
#
elf-pavlik ... eprodrom can you open issue on AS2.0 which would address need to making it 'make sense' from Social API point of view
#
elf-pavlik Arnaud: i worry that we define this issue very vague, how can we tell: we're done with it - resolved
#
sandro "Does AS2 Need to be different to support the API?"
#
elf-pavlik tantek: how about defining exit conditions for that issue
#
tantek ack sandro
#
Zakim sees no one on the speaker queue
#
elf-pavlik eprodrom: even having FWPD could work as exit condition
#
elf-pavlik sandro: "what changes does the spec need to support the API"
#
Arnaud do we know which way the dependency is?
#
tantek Arnaud, it's bidirectional
#
elf-pavlik ... regarding broad review, do we need it in terms of Social API ?
#
Arnaud is the protocol depending on AS or the other way around?
#
elf-pavlik sandro, I didn't pay attention to AS2.0 since I didn't see it useful for any use cases we work with
#
elf-pavlik sandro: but after meeting in Paris I realized that AS2.0 can come useful as integral part of Social API
#
Arnaud are you saying they both are going to be a normative ref to the other?
#
elf-pavlik ... and we shouldn't go to CR without answering this answer
#
sandro No, I think it'll be a clean stack, Arnaud
#
Arnaud if that's the case, they will need to move on the REC track in sync
#
Arnaud so it's not bidirectional
#
elf-pavlik tantek: as a chair I would like to propose to publish a montly heartbeat for AS2.0 documents
#
elf-pavlik ... hopefully it will help somehow with getting more broad review (required for exiting CR)
#
elf-pavlik sandro: why monthly? what motivates such interval?
#
cwebber2 +1 for heartbeat
#
cwebber2 (-1 for heartbleed??)
#
sandro Arnaud, right, I expect *if* AS2 is designed properly, the API can just use it. But until we design the API, we wont know.
#
elf-pavlik tantek: with number of significant edits happening in last months, I see it useful to publish new drafts once a month
#
elf-pavlik jasnell: usually we haven't made major changes every month, if some requements come from Social API, most likely they will relate to vocabulary no so much to core
#
elf-pavlik ... i don't see why montly cycle would be benefitial for us
#
tantek q?
#
Zakim sees no one on the speaker queue
#
Arnaud I suggest we may ask ourselves the question of whether to publish every month
#
elf-pavlik ... definitely something shorter than 5-6 months but 1 month not really needed
#
Arnaud but not to commit to publishing
#
jasnell Arnaud: +1
#
tantek Topic: https://www.w3.org/wiki/Socialwg/2015-06-30#Social_API
#
eprodrom q+
#
Zakim sees eprodrom on the speaker queue
#
tantek in particular, review of http://w3c.github.io/web-annotation/protocol/wd/
#
tantek added by eprodrom
#
elf-pavlik eprodrom: last week we've heard about work from Web Annotations WG, including publishing FPWD
#
elf-pavlik ... draft bases on LDP and uses some of its constraints
#
elf-pavlik ... earlier we looked at aligning our work on Social API with Annotations WG
#
cwebber2 I read it
#
KevinMarks reading it now
#
csarven has very briefly
#
cwebber2 o/
#
elf-pavlik ... i have taken a look at this working draft, have anone else did?
#
csarven Moreso on the data model
#
cwebber2 but I don't know LDP super well
#
cwebber2 or solid :)
#
jasnell I've read it. Also had a good chat with Doug after the call last week
#
jasnell regarding positioning of AS2, annotations and the WG. I may be helping with authoring a Note that describes how AS2 and Annotations fit together.
#
elf-pavlik eprodrom: ... explaining his understanding of Web Annotaions ...
#
elf-pavlik ... a lot of this workflow fits with what we talked about in context of social interactions
#
cwebber2 eprodrom, was there following stuff on there? I didn't see that
#
elf-pavlik ... annotations draft doesn't specify workflow for updating annotations etc.
#
elf-pavlik ... same for following, reporting annotations back to original publisher etc.
#
csarven CRUDish
#
elf-pavlik ... if seems like very low level protocol which doesn't deal with high level stuff, feels in some ways similar to SoLiD proposal
#
elf-pavlik tantek: chairs note - mentioned draft says W3C Editor's Draft 18 June 2014 http://w3c.github.io/web-annotation/protocol/wd/
#
jasnell the date is confirmed to be a typo
#
jasnell confirmed that with doug last week
#
elf-pavlik ... not public Working Draft but only Editor's draft!
#
bigbluehat it's about to be published. It's been approved by the WG last Wednesday
#
cwebber2 https://github.com/w3c/web-annotation/commit/73934ad6b4c6d1ce9b69f7b08d5a6719859402e5
#
cwebber2 date has been being updated lately
#
cwebber2 in the git repo
#
wseltzer whispers that she has seen a publication request
#
elf-pavlik eprodrom: date is a typo, Doug pointed it out last time
#
csarven +1
#
csarven +1 to trackin git
#
tantek q?
#
Zakim sees eprodrom on the speaker queue
#
tantek ack eprodrom
#
Zakim sees no one on the speaker queue
#
tantek q?
#
Zakim sees no one on the speaker queue
#
elf-pavlik ... my question 1. does group wants to track this document?
#
cwebber2 it's interesting to me, I'm not sure what we would do with tracking it
#
elf-pavlik tantek: as chair i want to point out, there seams to be a lot of overlap in scope between this group and Web Annotations WG
#
jasnell q+
#
Zakim sees jasnell on the speaker queue
#
eprodrom Great diagram on http://www.w3.org/annotation/
#
elf-pavlik ... many annotations resemple interactions in activity streams model, replies comments etc.
#
tantek q?
#
Zakim sees jasnell on the speaker queue
#
tantek ack jasnell
#
Zakim sees no one on the speaker queue
#
bigbluehat "CfC: publish FPWD of Web Annotation Protocol; concluded with support" - https://lists.w3.org/Archives/Public/public-annotation/2015Jun/0293.html
#
cwebber2 +q
#
Zakim sees cwebber on the speaker queue
#
KevinMarks it makes more sense with http://www.w3.org/TR/annotation-model/ as well
#
elf-pavlik jasnell: after call last week with Doug we had good conversation about this overlap
#
elf-pavlik ... we came up with a plan to make a note on describing this overlap and what we could work on together
#
bigbluehat eprodrom doug's chart is fabulous...but it's also non-normative. notification and stuff hasn't been addressed as yet--which is why it's not noted in the protocol.
#
KevinMarks the WAA diagram is good, not sue hwo it relates to that container spec as that is about a silo
#
tantek PROPOSAL: Jasnell and shepazu co-edit a joint working group NOTE between Annotations WG and Social Web WG describing the overlap & relationship between AS2 and Web Annotation Protocol drafts.
#
ben_thatmustbeme should we wait until they reach FPWD?
#
KevinMarks the API does not show how to notify of an external annotation
#
elf-pavlik +1
#
jasnell +1
#
melvster +1
#
tantek +1
#
ben_thatmustbeme +1
#
sandro +1
#
cwebber2 +1 but
#
cwebber2 I do have a comment / question
#
eprodrom -1
#
tantek ack cwebber2
#
Zakim sees cwebber on the speaker queue
#
csarven *notes*
#
KevinMarks q+
#
Zakim sees cwebber, KevinMarks on the speaker queue
#
ben_thatmustbeme has figured out all by 3 of the callers in webex :P
#
tantek ack cwebber
#
Zakim sees KevinMarks on the speaker queue
#
eprodrom q+
#
Zakim sees KevinMarks, eprodrom on the speaker queue
#
elf-pavlik sorry missed you cwebber2 can you please minute it?
#
tantek ack eprodrom
#
Zakim sees KevinMarks on the speaker queue
#
cwebber2 elf-pavlik, sure
#
elf-pavlik tantek: can you elaborate your -1
#
eprodrom http://www.w3.org/TR/annotation-model/
#
csarven Makes more sense with SoLiD, no?
#
Arnaud ben_thatmustbeme, unfortunately this info doesn't get shared, it's only local to your client
#
cwebber2 cwebber2: I see similarities and it's interesting but I'm not sure how this extends to the full social API, would be good to map to user stories
#
elf-pavlik eprodrom: i feel like AS2.0 and Web Annotation model are more equivalent and better to dicsuss them rather than possibly paralell Social API and Web Annotations protocol
#
elf-pavlik thanks cwebber2!
#
elf-pavlik jasnell: in the note we will focus on relation between data model and fundamental use cases
#
cwebber2 elf-pavlik: no prob
#
cwebber2 oops
#
tantek q?
#
Zakim sees KevinMarks on the speaker queue
#
cwebber2 dang colon vs comma
#
elf-pavlik s/elf-pavlik: no prob//
#
elf-pavlik s/oops//
#
cwebber2 :)
#
elf-pavlik :)
#
KevinMarks franjkly, http://www.w3.org/annotation/diagrams/annotation-architecture.svg is much better than either the data model or API doc
#
eprodrom +1
#
ben_thatmustbeme cwebber2 it doesn't mess up minutes for this WG, others might have an issue though apparently,
#
cwebber2 ben_thatmustbeme, good to know, though I seem to get consistent flak for making that mistake so :)
#
elf-pavlik sorry folks, i get bit lost here :(
#
ben_thatmustbeme cwebber2 me too, i just ignore it
#
elf-pavlik jasnell: focus on compatibility of AS2.0 model and Web Annotations model
#
ben_thatmustbeme cwebber2 its default behaviour in pretty much EVERY IRC client, so its rediculous to harp on it when it makes no difference to this group
#
elf-pavlik tantek: do you want to clarify the proposal? i think we need clarification from you for the end scope
#
ben_thatmustbeme elf-pavlik, remember as scribe you can stop conversation if you need to
#
elf-pavlik jasnell: note will focus on relation of AS2.0 and work of Web Annotations WG
#
elf-pavlik i lost audio!!!!
#
elf-pavlik can someone elese scribe for a while
#
eprodrom I can do it
#
eprodrom scribenick: eprodrom
#
ben_thatmustbeme was just about to jump in
#
eprodrom scribe: eprodrom
#
tantek PROPOSAL: Jasnell and shepazu co-edit a joint working group NOTE between Annotations WG and Social Web WG describing the overlap & relationship between AS2 and both Web Annotation Data Model and what they're trying to accomplish with Web Annotation Protocol.
#
jasnell +1
#
eprodrom +1
#
ben_thatmustbeme before that, why don't you clear the q
#
tantek q?
#
Zakim sees KevinMarks on the speaker queue
#
tantek q?
#
Zakim sees KevinMarks on the speaker queue
#
tantek ack KevinMarks
#
Zakim sees no one on the speaker queue
#
eprodrom KevinMarks: It's doesn't seem to conform to the architecture that we all liked
#
eprodrom KevinMarks: Only shows ways to post annotations onto a particular site, while the architecture shows doing it across multiple sites
#
eprodrom KevinMarks: All the examples are on the same domain
#
eprodrom KevinMarks: Which for us, we're looking at interactions across multiple sites
#
eprodrom KevinMarks: So, that's an important issue to discuss with Annotations WG
#
eprodrom tantek: Is that a blocking issue for this note?
#
eprodrom KevinMarks: It's something worth review. It's fairly key to our model of how we think about this.
#
eprodrom tantek: In our microformats examples we try to use different domains.
#
eprodrom tantek: this sounds like something we need to discuss in this note.
#
eprodrom KevinMarks: Will write a note to put up and point to.
#
tantek +1
#
ben_thatmustbeme +1
#
cwebber2 +1
#
KevinMarks +1
#
eprodrom +1
#
tantek RESOLVED: Jasnell and shepazu co-edit a joint working group NOTE between Annotations WG and Social Web WG describing the overlap & relationship between AS2 and both Web Annotation Data Model and what they're trying to accomplish with Web Annotation Protocol.
#
elf-pavlik iPad i used for webex client overheated :( i can't join audio any more... eprodrom please continue scribing if you can and i will add minutes to the wiki
#
Loqi aww, cheer up
#
eprodrom elf-pavlik, will do
#
tantek topic: https://www.w3.org/wiki/Socialwg/2015-06-30#Tracking_of_Actions_and_Issues
#
elf-pavlik eprodrom++
#
Loqi eprodrom has 18 karma
#
tantek no raised: http://www.w3.org/Social/track/issues/raised
#
eprodrom tantek: In our chairs discussion, we decided to move handling of issues and actions to the end of the meeting.
#
tantek no pending either: http://www.w3.org/Social/track/issues/pendingreview
#
eprodrom tantek: No raised issues
#
eprodrom tantek: no pending issues.
#
jasnell there are four change proposals in github
#
eprodrom tantek: There are no pending action items.
#
tantek no pending actions either: http://www.w3.org/Social/track/actions/pendingreview
#
jasnell q+
#
Zakim sees jasnell on the speaker queue
#
cwebber2 nope
#
eprodrom tantek: any reports on actions and issues that the group should know about?
#
tantek ack jasnell
#
Zakim sees no one on the speaker queue
#
eprodrom jasnell: Looking at Github issue tracker, there are 9 open issues
#
jasnell https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134
#
jasnell https://github.com/jasnell/w3c-socialwg-activitystreams/issues/133
#
jasnell https://github.com/jasnell/w3c-socialwg-activitystreams/issues/131
#
jasnell https://github.com/jasnell/w3c-socialwg-activitystreams/issues/86
#
cwebber2 I have to go, sorry
#
cwebber2 will follow the logs when I get back
#
tantek ACTION everyone review https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134 and /133 and /131 and /86 (in any order, independent of each other) by next week for in-call discussion.
#
trackbot is creating a new ACTION.
#
trackbot Error finding 'everyone'. You can review and register nicknames at <http://www.w3.org/Social/track/users>.
#
eprodrom tantek: any outstanding business?
#
eprodrom q+
#
Zakim sees eprodrom on the speaker queue
#
tantek ack eprodrom
#
Zakim sees no one on the speaker queue
#
eprodrom unmuting 1s
#
tantek note, next week's telcon will be on schedule, Tuesday July 7th, and Arnaud will be chairing
#
ben_thatmustbeme rhiaro is not on the call
#
eprodrom eprodrom: Upcoming milestones in Social API brainstorming?
#
aaronpk q+
#
Zakim sees aaronpk on the speaker queue
#
tantek ack Arnaud
#
Zakim sees aaronpk on the speaker queue
#
tantek ack aaronpk
#
Zakim sees no one on the speaker queue
#
rhiaro huh?
#
elf-pavlik sandro, i mentioned you originally in issue 134, i thought you may have some feedback on "@vocab": "_:" hackhttps://github.com/jasnell/w3c-socialwg-activitystreams/issues/134
#
eprodrom aaronpk: Can't speak for Amy and Tsyesika, have had other priorities in last couple of weeks, will have time in next couple of weeks
#
ben_thatmustbeme rhiaro there was a question of updates on brainstorming doc
#
eprodrom aaronpk++
#
Loqi aaronpk has 866 karma
#
rhiaro What aaronpk said!
#
tantek https://indiewebcamp.com/2015
#
eprodrom tantek: Indiewebcamp 2015 event happening July 12 in Brighton, Portland, and possibly NYC
#
rhiaro 26-7 in Edinburgh :)
#
eprodrom tantek: Open to anyone who has or wants to have their own site
#
eprodrom tantek: Signup links are on shared URL
#
eprodrom Cool!
#
eprodrom tantek: Ending the call
#
eprodrom elf-pavlik, will you edit the minutes?
#
elf-pavlik eprodrom, yes!
#
tantek elf-pavlik++ for minuting
#
Loqi elf-pavlik has 29 karma
#
tantek eprodrom++ for minuting when elf-pavlik's phone connection dropped!
#
Loqi eprodrom has 19 karma
#
elf-pavlik eprodrom++
#
Loqi eprodrom has 20 karma
#
tantek trackbot, end meeting
#
trackbot Zakim, list attendees
#
trackbot is ending a teleconference.
#
Zakim sorry, trackbot, I don't know what conference this is
#
trackbot RRSAgent, please draft minutes
#
RRSAgent I have made the request to generate http://www.w3.org/2015/06/30-social-minutes.html trackbot
#
trackbot RRSAgent, bye
#
tantek Zakim, this is SOCL
#
Zakim tantek, I see T&S_SOCWG()1:00PM in the schedule but not yet started. Perhaps you mean "this will be SOCL".
#
RRSAgent I see no action items
#
tantek Zakim, this was SOCL
#
Zakim I don't understand 'this was SOCL', tantek
#
aaronpk poor zakim
#
tantek Zakim is losing it
#
tantek thanks everyone!
#
eprodrom jasnell: dang it
#
eprodrom I just accidentally committed another spelling change to the vocab document directly
#
eprodrom If you don't mind I'm not going to go back and revert and do a PR
#
eprodrom I think I'm just going to stop using the in-page editor
#
eprodrom The default is "commit directly"
#
elf-pavlik eprodrom, IMO we also get quite some emails noise to the list with so many small PRs
#
tantek eprodrom or we could propose making you a co-editor :)
#
elf-pavlik today's minutes available on: https://www.w3.org/wiki/Socialwg/2015-06-30-minutes
#
tantek elf-pavlik++
#
Loqi elf-pavlik has 30 karma
#
Loqi Pelf made 1 edit to [[Socialwg/2015-06-30-minutes]] https://www.w3.org/wiki/index.php?diff=84900&oldid=0
#
ben_thatmustbeme eprodrom: yeah, the in-page editor means an immediate PR which means a message to the mailing list of every change, but makes for pretty colors https://github.com/jasnell/w3c-socialwg-activitystreams/network
#
eprodrom I try to make atomic changes to leave it up to the editor to accept
#
eprodrom Sorry about the mail though
#
elf-pavlik eprodrom, you can make one PR with multiple commits which still allow cherry-picking
#
eprodrom elf-pavlik: Thanks, but I'm going to stick to the way I'm doing it right now
#
elf-pavlik eprodrom, as you wish
#
ben_thatmustbeme or they can always just pull the PR into a branch and modify edits there
#
elf-pavlik jasnell, do you have any preference?
#
elf-pavlik aaronpk, do you have link to example of "p3k has support for private posts by specifying a list of the domains of people authorized to view a private post. " ?
#
elf-pavlik kylewm, same for you :) "Red Wind has rudimentary support for private posts based on a simple list of the domains of people authorized to view a private post."
#
jasnell eprodrom: no worries
#
ben_thatmustbeme elf-pavlik: thats sort of difficult to give an example of publicly, since they are private
#
ben_thatmustbeme i have successfully viewed aaronpk's private posts though
#
elf-pavlik example of ACL file not the post itself
#
tantek ben_thatmustbeme: perhaps you could post a private note to elf-pavlik :)
#
ben_thatmustbeme tantek: i don't post private notes right now
#
ben_thatmustbeme its on my to-do list after i get contacts list working
#
ben_thatmustbeme elf-pavlik: i'm pretty sure he doesn't use an ACL file, its specified entirely internally
#
melvster ben_thatmustbeme: deiu just finished writing a contacts list app ...
#
melvster elf-pavlik: re your earlier question: I got a reply on the list: "One problem is that you can't use blank nodes as predicates (unless you are using some generalized extension of RDF) . "
#
elf-pavlik tantek, i made quick prototype https://elf-pavlik.github.io/unmpc/ you can use it to post to group spaces https://phubble.tuxed.net/w3c-social/ (public) and https://phubble.tuxed.net/alice-b-day/ (secret) both using same file to define ACL https://github.com/w3c-social/social-web/blob/master/phubble-acl.json
#
elf-pavlik i need to fix parsing microformats from secret spaces HTML (small issue with the way i select nodes, but you can see in dev console HTML coming if client app makes request with Bearer Token)
#
Loqi Tantekelik made 1 edit to [[Socialwg/2015-06-30]] https://www.w3.org/wiki/index.php?diff=84901&oldid=84899
#
melvster that annotations spec sure looks interesting
#
elf-pavlik if i can't see private note using reader of my preference i don't see it very somooth experience to run all over the web and fiddle with all kind of UIs just to collect my messages :(
#
Loqi it'll be ok
#
elf-pavlik unmpc just uses Bearer Token for READ just the same way as for WRITE using micropub endpoint of group wall
#
eprodrom jasnell: I've gotten about as far as I'm going to go today
#
eprodrom So feel free to proceed without further input from me
#
jasnell eprodrom: thank you very much. it's definitely appreciated
#
eprodrom NP
#
eprodrom I've got 2 pages of notes and issues
#
jasnell eprodrom++
#
eprodrom But I'll put them into Github over the next couple of days
#
eprodrom ha
#
eprodrom don't plus plus me until you see the issues
#
eprodrom They could be awful
#
jasnell :-)
#
elf-pavlik aaronpk, kylewm do you use webmention to notify people about private posts ?
#
melvster jasnell: do you have thoughts on the default AS2 @vocab being _: ?
#
melvster I had a crazy idea of using urn: there ... any thoughts?
#
aaronpk elf-pavlik: i want to, but that really only works if the receiver knows to re-try fetching the post with auth
#
aaronpk for now i've been sending the link via other channels, like IRC or email
#
elf-pavlik aaronpk, i get this web mention to my reader which i can use to get bearer token for read private note, then it could work i guess ...
#
jasnell melvster: I'm perfectly fine with the default mapping to _:
#
melvster jasnell: ok thanks, that's good to know
#
jasnell I understand it could lead to some issues when looking at the world through RDF lenses, but it works well in terms of the JSON-LD processing
#
melvster jasnell: there is that, but, althought that wasnt my primary thought process, I was considering an unknown variable to be the same as a "name", rather than, an unnamed thing (ie bnode)
#
melvster so "key" to me sounds more like urn:key than _:key because the programmer would have called it 'key' for a reason
#
elf-pavlik jasnell, bblfish suggested tests with mergning data from AS2.0 orgiginating from different publishers, i would like to add example from issue 134 to such test https://github.com/jasnell/w3c-socialwg-activitystreams/issues/134#issuecomment-108255400
#
melvster with _: essentially it's taking away the name
#
melvster i just wonder in the world of web2.0 and social json, maybe there's a lot of json named properties that were given that name in a non opaque way
#
jasnell well, JSON-LD expansion and compaction preserves the name even if it's _:key, it's only when you perform additional processing beyond that that the name can get dropped
#
elf-pavlik IMO system should treat _:foo != _:foo while ex:foo == ex:foo ... not sure how consumer can do this without applying JSON-LD context to flattened data?
#
elf-pavlik _:foo != _:foo if they come from different documents
#
jasnell the key issue for me, however, is just that data is not dropped on the floor and that properties and object types that are not defined in a context are put into a non-namespaced bucket
#
jasnell elf: that's fine that _:foo != _:foo if they come from different documents
#
melvster jasnell: urn:foo is namespaced tho
#
melvster urn just says 'this is a name, we dont know anything more'
#
jasnell publishers need to be aware that using terms that are not defined in a context can lead to inconsistent behavior
#
jasnell that's why the spec warns against it
#
elf-pavlik how can consumer not applying JSON-LD context know which properties can consider equal and which not?
#
jasnell if those properties are not defined in a context, they cannot. two different implementers may use the same term to represent different concepts
#
jasnell so if the publisher is concerned about that, they need to make sure the terms they use are defined in a context
#
elf-pavlik only http://www.w3.org/TR/activitystreams-vocabulary/ can consider equal, for other it requers information from JSON-LD context
#
jasnell effectively, any term used that is not in a context is private use with no interop guarantees
#
melvster well its nice in a way, more anything that understands _: will also understand urn: so it's giving more power to the processors
#
elf-pavlik jasnell, what do you think about adding warning: "If you don't assign full URIs in JSON-LD context for custom terms that you use, consumers can NOT integrate your data with data from other sources" ?
#
jasnell mapping to urn: would have the same basic effect, yes, plus it would play nicer if someone wanted to do full RDF processing
#
jasnell but, in the end, the result is the same
#
jasnell if we map undefined terms to "urn:", then "urn:key" in one document != "urn:key" in another document
#
jasnell elf-pavlik: the spec already discusses the need for adding extensions to the context
#
elf-pavlik hmmmm.... maybe at least validator can warn when it detects falling back to _: ?
#
jasnell elf-pavlik: that would be beneficial
#
elf-pavlik just so people know that it comes with some draw backs
#
elf-pavlik should i create issue in https://github.com/w3c/activitystreams-testing/issues ?
#
jasnell if you'd like
#
melvster jasnell: i may be wrong there, but my understanding is that bnodes generally should be thought of as blank, and in most cases, not suitable for comparisons (I might be mistaken tho)
#
jasnell melvster: right, and that's exactly the point. If you want a term that is suitable for comparisons, make sure it's defined in a context
#
elf-pavlik done: https://github.com/w3c/activitystreams-testing/issues/4
#
jasnell e.g. {"@context": "http://www.w3.org/ns/activitystreams", "foo": 1} and {"@context": "http://www.w3.org/ns/activitystreams", "foo": "bar"}
#
jasnell the "foo" in each of these documents are not comparable. they each map to _:foo and _:foo, yes, but are still relative to their own documents
#
melvster jasnell: i get it, its interesting, tho lets say one system uses "longditude" and another "longditude" is there value of software introspecting, or of treating them as opaque, each route has dangers
#
elf-pavlik melvster, do you propose falling back to urn:foo: and urn:foo: in case of any use of not mapped foo property ?
#
jasnell implementations could apply some form of heuristics to attempt to reconcile the two but that would be going beyond the spec. It's possible, yes, but not a requirement
#
melvster elf-pavlik: at this point it was just an idea I had, and mainly looking for feedback saying, 'thats a bad idea because ...' ... I wonder ive not see it done before, so not proposing it, but i thought it may be an interesting discussion point, maybe something missed
#
elf-pavlik melvster, IMO mapping two different foo properties to same urn:foo could make things even worst than how i see this _: trick
#
melvster why?
#
jasnell elf-pavlik: yes
#
elf-pavlik consumer would end up with properties which have same URI but got used with compleately different intention and meaning
#
jasnell melvster: because it gives a false perception that the terms can be compared
#
melvster but dont you have that already with JSON?
#
elf-pavlik catching _:foo != _:foo when they come from two documents sounds doable, if you map it to same URI you lost the distinction
#
melvster isnt it just keeping the status quo?
#
jasnell if you're doing JSON only processing, yes.
#
melvster yeah it's tricky i can see both sides
#
jasnell if you're doing JSON-LD processing, then no
#
jasnell the goal here is to find a practical balance, not to make it perfect :-)
#
melvster i wonder tho should middleware assume urn:foo (from doc1) === urn:foo (from doc2)
#
jasnell melvster: it would have to
#
elf-pavlik jasnell, IMO we neede more https://www.w3.org/wiki/Socialwg/Deployments to get some real hands one experience
#
elf-pavlik s/one/on/
#
jasnell consider... {"@context": ["http://www.w3.org/ns/activitystreams", {"bar": "urn:bar"}], "foo": 1, "bar": "a"} and {"@context": ["http://www.w3.org/ns/activitystreams", {"bar": "urn:bar"}], "foo": "z", "bar": "a"}
#
elf-pavlik BTW +1 http://jasnell.github.io/w3c-socialwg-activitystreams/activitystreams2-vocabulary.html#dfn-instrument
#
melvster i cant actually think of an example when you'd use urn:foo in a comparison other than say var_dump or print_r or stringify
#
jasnell when expanded, doc A would have urn:bar="a", urn:foo=1 and doc B would have urn:bar"a" and urn:foo="z".
#
melvster LD stuff will only generally be dealing with http: uris
#
jasnell which again, gives the false sense that the urn:foo in each doc is comparable, when they are not
#
jasnell generally, but not guaranteed
#
jasnell it can be *any* URI, including any urn
#
jasnell using _: makes it absolutely unambiguous
#
melvster sure but the business logic will normally not do anything with urn:
#
melvster but rather, with http:
#
melvster or sometimes mailto:
#
jasnell can you guarantee that?
#
melvster i cant think of where you'd actually fault on this
#
melvster it would just be in the bucket 'unknown predicate/key/attribute'
#
melvster if i put urn:foo in an html document nothing with break because no one knows what to do with it
#
melvster it's just saying 'someone named this uri foo'
#
melvster perhaps im overthinking it!
#
melvster your comments make a lot of sense to me tho, appreciate you explaining
#
jasnell in practice, if everyone just used http: then I would agree, but this needs to work independently of the url protocol scheme that is selected
#
jasnell _: allows us to do that, without having to overthink it or even care
#
melvster jasnell: sure but whenever you get a URI, you're always going to introspect on the scheme/protocol to know what to do with it next
#
jasnell not necessarily. @type's can be completely opaque identifiers without having to introspect
#
melvster good point
#
melvster jasnell++
#
Loqi jasnell has 23 karma
#
jasnell e.g. if I see @type:"urn:example:Foo", I don't have to introspect it to know that it's a different thing than "http://example.org/Foo"
#
jasnell now, if I want to know more about "urn:example:Foo" then that particular identifier isn't going to help me very much, but it's still perfectly legal
#
jasnell I'm perfectly fine publishing a non-normative best practice that says "You really really really ought to be using http(s) url's as identifiers" but it's not necessary for us to *require* it. Let implementers do what they want
#
melvster jasnell: i tend to always check where something came from before doing comparisons ... you should always have a source for your data
#
melvster ie working with quads rather than working with triples
#
jasnell as a best practice, yes
#
melvster jasnell: id consider it essential, none of my apps would work without it, it's more than a nice to have for me
#
melvster there are times you can get away with not looking
#
melvster but in many cases it's really important
#
jasnell I agree
#
jasnell it's particularly important for interop around extensions
#
melvster yes
#
jasnell which is why if a publisher really wants interop, they need to define their terms
#
melvster yes, in an ideal world
#
jasnell if they don't define the terms in the @context, they're on their own
#
jasnell and if they choose to use urn schemes that do not support introspection and discovery, they're on their own also
#
melvster jasnell: so another example ... let's say your software came across urn:James ... what would it make of it?
#
melvster and would another urn:James be problematic?
#
jasnell an implementation would be required to treat urn:James in one document as comparable to urn:James is another
#
jasnell but that doesn't mean the implementation would be required to actually do anything with that information
#
melvster but all it's saying is 'both these entities have the name James' -- right?
#
jasnell yes
#
melvster they could be two different James'
#
jasnell are you familiar with Symbol in ES6?
#
melvster jasnell: ive skimmed it before, but not as familiar as id like to be
#
jasnell ok. Symbol("James") !== Symbol("James")
#
melvster ah right, yes
#
jasnell but Symbol.for("James") === Symbol.for("James")
#
melvster interesting
#
jasnell _:James is equivalent to Symbol("James")
#
jasnell urn:James is equivalent to Symbol.for("James")
#
melvster i guess we can solve this in AS6, then? :P
#
jasnell :-D
#
jasnell melvster++
#
Loqi melvster has 15 karma
#
melvster got to go ... thanks for the chat, v helpful
#
jasnell of course!
#
jasnell ttyl melvster
jdp23 joined the channel
#
Zakim excuses himself; his presence no longer seems to be needed
bblfish_, KevinMarks, shepazu, Arnaud and jasnell joined the channel