15:02:15 <RRSAgent> RRSAgent has joined #json-ld
15:02:15 <RRSAgent> logging to https://www.w3.org/2018/09/21-json-ld-irc
15:02:21 <azaroth> zakim, this is JSON-LD
15:02:21 <Zakim> got it, azaroth
15:10:36 <uskudarli> uskudarli has joined #json-ld
15:47:44 <ajs6f> ajs6f has joined #json-ld
15:52:03 <azaroth> Chair: azaroth
15:52:09 <azaroth> rrsagent, set minutes public
15:52:09 <RRSAgent> I'm logging. I don't understand 'set minutes public', azaroth.  Try /msg RRSAgent help
15:53:11 <azaroth> Agenda: https://lists.w3.org/Archives/Public/public-json-ld-wg/2018Sep/0015.html
15:53:15 <azaroth> azaroth has changed the topic to: Agenda: https://lists.w3.org/Archives/Public/public-json-ld-wg/2018Sep/0015.html
15:53:33 <azaroth> present+ Rob_Sanderson
15:53:44 <azaroth> Date: 2018-09-21
15:53:49 <azaroth> rrsagent, set log public
15:53:55 <azaroth> regrets+ ivan
15:53:58 <azaroth> regrets+ timCole
15:54:03 <azaroth> regrets+ simon
15:54:13 <azaroth> Meeting: JSON-LD Working Group Telco
15:55:03 <azaroth> regrets- simon
15:55:11 <azaroth> regrets+ simonstey
15:57:04 <gkellogg> gkellogg has joined #json-ld
15:58:12 <jeff_mixter> jeff_mixter has joined #json-ld
15:58:17 <jeff_mixter> present+
15:58:33 <gkellogg> present+ Gregg_Kellogg
15:59:05 <workergnome> workergnome has joined #json-ld
15:59:09 <workergnome> present+
16:00:52 <ajs6f> present+
16:01:09 <azaroth> TOPIC: Scribe Selection
16:01:50 <azaroth> zakim, who is here?
16:01:50 <Zakim> Present: Rob_Sanderson, jeff_mixter, Gregg_Kellogg, workergnome, ajs6f
16:01:51 <hsolbrig> hsolbrig has joined #json-ld
16:01:52 <Zakim> On IRC I see workergnome, jeff_mixter, gkellogg, ajs6f, uskudarli, RRSAgent, Zakim, azaroth, dlongley, dlehn, bigbluehat, fr33domlover, ChristopherA
16:01:56 <hsolbrig> present+ hsolbrig
16:02:57 <azaroth> scribenick: ajs6f
16:03:13 <azaroth> TOPIC: Approve minutes of previous call
16:03:31 <azaroth> PROPOSAL: Approve minutes of previous call: https://www.w3.org/2018/json-ld-wg/Meetings/Minutes/2018/2018-09-14-json-ld
16:03:33 <ajs6f> +1
16:03:34 <azaroth> +1
16:03:37 <hsolbrig> +1
16:03:38 <gkellogg> +1
16:03:38 <jeff_mixter> +1
16:04:01 <workergnome> +1
16:04:03 <azaroth> RESOLVED: Approve minutes of previous call: https://www.w3.org/2018/json-ld-wg/Meetings/Minutes/2018/2018-09-14-json-ld
16:04:34 <azaroth> TOPIC: Announcements / Reminders
16:05:10 <dlehn> present+ David_Lehn
16:05:10 <bigbluehat> present+ Benjamin_Young
16:05:18 <azaroth> TOPIC: Issues
16:05:32 <azaroth> SUBTOPIC: Framing as a verb
16:05:42 <azaroth> link: https://github.com/w3c/json-ld-framing/issues/13
16:06:11 <ajs6f> azaroth: OP notes that "framing" isn't "surrounding with a structure"
16:06:16 <ajs6f> ... so why call it framing?
16:06:33 <ajs6f> ... but that ship has sailed: the term is in use
16:06:48 <azaroth> q?
16:06:51 <bigbluehat> q+
16:07:11 <azaroth> ack bigbluehat
16:07:21 <ajs6f> bigbluehat: q about short names.
16:07:45 <ajs6f> ... "JSON-LD 1.1" is the name of one of our docs, version issues later might make that make no sense
16:08:05 <ajs6f> azaroth: Short names must be approved by W3C mgmnt, because they serve as "namespaces"
16:08:13 <ajs6f> ... but working drafts do not
16:08:23 <gkellogg> q+
16:08:28 <ajs6f> ... changing the short name is more involved than creating a new version
16:08:39 <azaroth> ack gkellogg
16:08:58 <ajs6f> gkellogg: it's been called "JSON-LD Framing" from the beginning
16:09:15 <ajs6f> gkellogg: wasn't part of RDF 1.1, but has been known as "framing" since then
16:09:24 <ajs6f> ... dlehn may know more
16:09:31 <ajs6f> ... but the ship really has sailed
16:09:44 <azaroth> PROPOSAL:  Too late to change the name at this stage, close framing#13 won't fix.
16:09:48 <bigbluehat> +1
16:09:48 <azaroth> +1
16:09:50 <ajs6f> ... if this were new, maybe we could change names, but not now
16:09:54 <gkellogg> +1
16:09:58 <hsolbrig> +1
16:10:11 <bigbluehat> q+
16:10:12 <dlehn> +1
16:10:58 <ajs6f> +1
16:11:03 <jeff_mixter> +1
16:11:08 <azaroth> ack bigbluehat
16:11:10 <workergnome> +1
16:11:15 <azaroth> RESOLVED:  Too late to change the name at this stage, close framing#13 won't fix.
16:11:46 <azaroth> SUBTOPIC: subtitle for framing
16:11:48 <azaroth> link: https://github.com/w3c/json-ld-framing/issues/14
16:12:14 <ajs6f> gkellogg: the subtitle of this doc is the name as I just described: "JSON-LD"
16:13:05 <ajs6f> ... it's an extension to the API spec, but the subtitle doesn't make that clear
16:13:21 <ajs6f> +1
16:13:28 <workergnome> +1
16:13:30 <bigbluehat> +1
16:13:32 <hsolbrig> +1
16:13:33 <azaroth> +1
16:13:38 <jeff_mixter> +1
16:13:41 <gkellogg> +1
16:13:53 <dlehn> +1
16:13:59 <ajs6f> * hang on is that proposal prefixed by an asterix?
16:14:06 <azaroth> SUBTOPIC: Remove bnode naming requirement
16:14:20 <azaroth> link: ttps://github.com/w3c/json-ld-api/issues/25
16:14:33 <azaroth> s/ttps:/https:/
16:14:50 <ajs6f> gkellogg: creates some complications for testing, because you can't use bnode labels
16:15:00 <ajs6f> ... expansion does not rename bnodes
16:15:04 <ajs6f> ... flattening does
16:15:28 <ajs6f> ... but you can use either normalization or isomorphism to compare
16:16:17 <ajs6f> ... but for a structural comparision we compare triples and expand to check lists
16:16:24 <azaroth> q+ to ask about breaking-ness of change
16:16:26 <ajs6f> ... so that would include bnode names
16:16:50 <ajs6f> ... so the only thing left is to flatten, turn into triples, and compare to see that the graphs are the _same_
16:17:15 <ajs6f> ... there was some discussion about using a bijection for comparison, but that made for a very complex test arrangement
16:17:34 <ajs6f> ... and we might not verify that the topology of the flattened doc is correct
16:17:59 <hsolbrig> q+
16:18:05 <azaroth> ack hsolbrig
16:18:16 <bigbluehat> q+
16:18:41 <ajs6f> hsolbrig: it concerns me to offer a bnode labelling algorithm-- it might "leak"
16:18:46 <gkellogg> https://json-ld.github.io/normalization/spec/
16:18:52 <ajs6f> gkellogg: dataset normalization offers this
16:19:46 <ajs6f> ... because you have to be able to offer _exactly_ the same quads
16:20:02 <ajs6f> hsolbrig: would support making bnode labelling non-normative
16:20:20 <ajs6f> ... understand the problem with testing, but more worried about becoming a "standard" for bnode labelling
16:20:26 <azaroth> ack azaroth
16:20:26 <Zakim> azaroth, you wanted to ask about breaking-ness of change
16:20:27 <ajs6f> gkellogg: it's already out there since 1.0
16:20:46 <ajs6f> ... so we might end up with a non-normative discussion instead
16:21:40 <ajs6f> azaroth: how breaking a change is this?
16:21:53 <dlehn> as someone who may have to deal with the testing implementation, i'm somewhat unclear on what problems may come up with this
16:21:54 <ajs6f> ... will we be breaking 1.0 systems
16:22:31 <ajs6f> gkellogg: depends on whether people depend on bnode names
16:22:40 <ajs6f> ... similar to bnodes for properties
16:23:19 <ajs6f> azaroth: what would be the requirement after this?
16:23:37 <ajs6f> gkellogg: thought it would be that bnode names must be unique and non-overlapping
16:24:10 <ajs6f> ... and if an algorithm can use our mapping, testing will be easier
16:24:10 <azaroth> q?
16:24:29 <ajs6f> ... q is: can we eliminate it, or mark it as obsolete?
16:24:49 <azaroth> ack bigbluehat
16:25:08 <ajs6f> bigbluehat: thinking from JSON perspective
16:25:11 <azaroth> q+ dlehn to ask what problems may come up with this
16:25:45 <ajs6f> ... suppose someone is using the algorithms to go from JSON, what effect does this have?
16:26:29 <ajs6f> ... these effects could be shocking to a JSON user
16:26:51 <ajs6f> gkellogg: these are unkowns
16:27:12 <ajs6f> ... smilar to what happens with most RDF parserse
16:27:35 <ajs6f> ... e.g. round trip from Turtle to NTriples, many people would expect bnode labels to be preserved
16:27:51 <jeff_mixter> q+
16:27:59 <ajs6f> ... but that's not guaranteed, just convenient
16:28:51 <azaroth> oooh, +1 to marking at risk
16:29:03 <ajs6f> ... might mark this "at-risk" to get more feedback
16:29:15 <azaroth> ack dlehn
16:29:15 <Zakim> dlehn, you wanted to ask what problems may come up with this
16:29:35 <ajs6f> dlehn: wanted to ask what problems may come up with this
16:29:52 <ajs6f> gkellogg: one type: people who rely on knowledge of the names generated
16:30:04 <ajs6f> ... which is just a really bad thing to do, but they might have done it
16:30:19 <dlehn> mostly my questions are how much pain this could cause for the test runner implementation ;)
16:30:20 <ajs6f> ... then there's ordering.
16:30:40 <ajs6f> ... part of the purpose of this algo is to preserve document order
16:31:05 <ajs6f> ... but if bnode names are different, could reorder output
16:31:56 <ajs6f> ... could use an unorder comparison for tests
16:32:17 <ajs6f> ... or could reduce each to triples and then again, normalization or compare by isomorphism
16:32:34 <dlehn> sounds like this will need an early set of tests just to test that the test runner comparison code works
16:32:41 <ajs6f> ... would require more testing and discussion
16:32:52 <ajs6f> ... will be a pain for devs
16:33:03 <azaroth> ack jeff_mixter
16:33:04 <ajs6f> ... but devs could continue to use the informativ algos
16:33:19 <ajs6f> jeff_mixter: more of a philosophical apporach
16:33:41 <ajs6f> ... agree with "data-first" approach, think of bnodes as wildcards
16:33:50 <ajs6f> ... sad if devs have relied on naming
16:33:58 <azaroth> q?
16:34:04 <ajs6f> ... but it's the problem of parsing, not the data
16:34:59 <ajs6f> azaroth: some might argue that we broke their nice little code
16:35:05 <gkellogg> q+
16:35:10 <ajs6f> ... if they're sad, they can respond to the "at risk" status
16:35:14 <azaroth> ack gkellogg
16:35:16 <ajs6f> ... and if they don't we go ahead
16:35:37 <ajs6f> gkellogg: we can make this more visible in the next WD, with a blog post
16:35:39 <azaroth> +1 to blog post on this
16:35:50 <jeff_mixter> +1 to post
16:35:57 <bigbluehat> +1 to blog post
16:36:50 <workergnome> +1
16:36:52 <azaroth> PROPOSAL: Mark normative blank node naming algorithm at risk in next WD plus blog post about this issue and obsolence of blank node as property ; if no feedback than proceed in subsequent WD
16:36:56 <gkellogg> +1
16:36:58 <workergnome> +1
16:36:59 <azaroth> +1
16:36:59 <hsolbrig> +1
16:37:00 <jeff_mixter> +1
16:37:01 <bigbluehat> +1
16:37:05 <dlehn> +1
16:37:08 <ajs6f> +1
16:37:18 <azaroth> RESOLVED: Mark normative blank node naming algorithm at risk in next WD plus blog post about this issue and obsolence of blank node as property ; if no feedback than proceed in subsequent WD
16:37:34 <azaroth> SUBTOPIC: Base of an embedded JSON-LD block
16:37:47 <azaroth> link: https://github.com/w3c/json-ld-syntax/issues/23
16:38:19 <ajs6f> azaroth: we don't normatively spec the base URI for a JSON-LD block
16:38:21 <ajs6f> ... inside HTML
16:38:27 <ajs6f> ... could be several choices
16:38:44 <ajs6f> ... some of which could be further dependenet on other parts of the HTML doc
16:39:14 <ajs6f> ... the ticket has discussion for other problems with embedded JSON-LD, but let's stick to base URI for this issue
16:39:23 <azaroth> q?
16:39:26 <bigbluehat> q+
16:39:28 <ajs6f> ... base is relatively clear
16:39:37 <ajs6f> gkellogg: it's the principle of least surprise
16:39:52 <ajs6f> ... if I load an HTML doc and it include JSON-LD w/ relative URIs
16:39:55 <ajs6f> ... they must be resolved
16:40:01 <ajs6f> ... make sense to use the doc base
16:40:16 <ajs6f> ... but what if I change the doc base using base head element
16:40:24 <workergnome> q+
16:40:26 <ajs6f> ... used in RDFa
16:40:45 <ajs6f> ... would be odd if JSON-LD and RDFa in same page had diff bas
16:40:50 <ajs6f> s/bas/base/
16:41:06 <ajs6f> ... could be a pain, but it is what the DOM uses
16:41:29 <ajs6f> ... if we do something different, that would be even more surprising
16:42:28 <azaroth> ack bigbluehat
16:42:38 <ajs6f> bigbluehat: all kinds of concerns about this
16:42:58 <ajs6f> ... about using data from non-Javascript mimetype surroundings
16:43:30 <ajs6f> ... yes, you can have all kinds of info about a script tag, but I wouldn't expect it to change the text extracted from that script tag
16:43:48 <azaroth> q?
16:43:55 <ajs6f> ... requiring interplay between the DOM and what's in the script tag is a dangerous precedent and would have a lot of overhead
16:44:18 <ajs6f> ... would expect the base of the JSON-LD would be the same as if it were raw, sent by itself
16:44:52 <azaroth> ack workergnome
16:45:21 <ajs6f> dlehn: while trying to generate framed json-ld which is not publshed
16:45:25 <ajs6f> ... and therefore have no base
16:45:32 <ajs6f> ... I've had to invent one to get the algos to work
16:45:41 <bigbluehat> q+
16:45:50 <ajs6f> ... our docs seem to assume that the JSON-LD exists "at a location"
16:46:01 <azaroth> ack bigbluehat
16:46:28 <ajs6f> bigbluehat: that affects Linked Data all over
16:46:33 <azaroth> q?
16:46:49 <ajs6f> ... at WIley we do a lot of RDFa and JSON together
16:47:00 <gkellogg> q+
16:47:02 <ajs6f> ... we use a hash
16:47:27 <azaroth> ack gkellogg
16:47:28 <ajs6f> workergnome: I've been creating a base, even just "example.com"
16:47:41 <ajs6f> gkellogg: has been an issue in the PLayground
16:47:55 <gkellogg> https://tools.ietf.org/html/rfc3986#section-5.1
16:48:30 <ajs6f> ... we want apps to have something to use
16:48:38 <ajs6f> ... in the absence of anything explicit
16:49:04 <ajs6f> ... if the mediatype is json-ld, we have a known way to establish a base with in the context
16:49:11 <ajs6f> ... but what about w/i a script tag
16:49:37 <ajs6f> ... we can say, it's based on the URI that was used to retrieve the document
16:49:46 <bigbluehat> q+ to point to some HTML5 stuff related to data-blocks
16:49:56 <azaroth> ack bigbluehat
16:49:56 <Zakim> bigbluehat, you wanted to point to some HTML5 stuff related to data-blocks
16:49:59 <ajs6f> ... there's doc base, head base, and xml:base
16:50:01 <bigbluehat> https://www.w3.org/TR/html5/semantics-scripting.html#data-block
16:50:42 <ajs6f> bigbluehat: in that paragraph HTML5 says nothing around the script tag can have any effect on the stuff inside it
16:50:49 <ajs6f> ... we would break that
16:51:05 <ajs6f> ... would make a species of special JSON-LD style HTML aprsers
16:51:12 <ajs6f> s/aprsers/parsers
16:51:25 <gkellogg> 1=
16:51:29 <ajs6f> ... moving the bar for consumption the wrong way-- towards harder
16:51:30 <gkellogg> q+
16:51:33 <azaroth> ack gkellogg
16:51:53 <ajs6f> gkellogg: excellent reference, can use that to solve this
16:52:08 <azaroth> +1
16:52:27 <ajs6f>  /me nice job, bigbluehat
16:52:47 <ajs6f> s/\/me nice job, bigbluehat//
16:53:10 <azaroth> PROPOSAL: When establishing the base URI, use the document base URL only and ignore all surrounding data such as xml:base
16:53:15 <gkellogg> +1
16:53:16 <azaroth> +1
16:53:22 <hsolbrig> +1
16:53:31 <workergnome> +1
16:53:35 <ajs6f> +1
16:53:52 <bigbluehat> +1
16:54:03 <hsolbrig> +1
16:54:04 <ajs6f> azaroth: e.g. if YAML came up with some internal way to change the base, this proposal would tell us to ignore that too
16:54:07 <dlehn> +1
16:54:13 <gkellogg> q+
16:54:19 <ajs6f> ... or any other serialization
16:54:25 <ajs6f> gkellogg: not sure we can be that broad
16:54:36 <ajs6f> ... this reference is only good for this mediatype
16:54:52 <bigbluehat> q+
16:54:56 <ajs6f> ... other mediatypes may have their own policies
16:55:11 <ajs6f> ... but in the absence of such specific guidance, we default to this stance from HTML5
16:55:12 <azaroth> RESOLVED: When establishing the base URI, use the document base URL only and ignore all surrounding data such as xml:base
16:55:26 <azaroth> Also +1 to Gregg's clarified clarification :D
16:55:31 <ajs6f> bigbluehat: but this is all within HTML5
16:55:57 <ajs6f> ... so that is the dictator we use as a conveyance
16:56:15 <ajs6f> ... but we've now nailed this to the wall succinctly
16:56:39 <ajs6f> azaroth: we can reuse this guidance for language and text direction
16:56:58 <azaroth> PROPOSAL: Language, text direction and similar are not affected by surrounding content, such as attributes on containing HTML elements
16:56:59 <gkellogg> q+
16:57:01 <azaroth> +1
16:57:02 <bigbluehat> +1
16:57:03 <azaroth> ack gkellogg
16:57:04 <ajs6f> +1
16:57:08 <bigbluehat> q-
16:57:19 <ajs6f> gkellogg: the wording here is all negative
16:57:44 <ajs6f> ... we could instead say "only this way is correct"
16:58:20 <azaroth> PROPOSAL: The only thing inherited by the JSON-LD processor is the document location established by HTTP for determining the base URI, unless the rules of the embedding media type dictate otherwise
16:58:29 <gkellogg> +1
16:58:30 <azaroth> +1
16:58:31 <dlehn> +1
16:58:34 <ajs6f> +1
16:58:37 <bigbluehat> +1
16:58:39 <hsolbrig> +1
16:58:54 <ajs6f> LASOPORP
16:59:22 <azaroth> RESOLVED: The only thing inherited by the JSON-LD processor is the document location established by HTTP for determining the base URI, unless the rules of the embedding media type dictate otherwise
16:59:44 <ajs6f> \me hat tip
17:00:11 <azaroth> TOPIC: Adjourn
17:00:15 <ajs6f> \ know what / am doing
17:01:16 <azaroth> rrsagent, publish log
17:01:16 <RRSAgent> I'm logging. I don't understand 'publish log', azaroth.  Try /msg RRSAgent help
17:01:40 <azaroth> rrsagent, draft minutes
17:01:40 <RRSAgent> I have made the request to generate https://www.w3.org/2018/09/21-json-ld-minutes.html azaroth
17:02:01 <azaroth> rrsagent, set log public
17:02:15 <azaroth> Yup, there it is :)
17:03:13 <azaroth> zakim, bye
17:03:13 <Zakim> leaving.  As of this point the attendees have been Rob_Sanderson, jeff_mixter, Gregg_Kellogg, workergnome, ajs6f, hsolbrig, David_Lehn, Benjamin_Young
17:03:13 <Zakim> Zakim has left #json-ld
17:03:17 <azaroth> rrsagent, bye
17:03:17 <RRSAgent> I see no action items