Meeting minutes
<pchampin> unfortunately I will only be able attend the 1st half-hour
ora: should scrutenize niklasl presentation today
ora already been some messages on mailing list
niklasl: split out parts on named graphs in separate slides
niklasl: only introduction remains currently - around 17 slides
ora: my slides are more or less complete; will refer to niklasl for details
<TallTed> s/niklasl: only/
ora: should likely not copy niklasl's slides into my deck
<TallTed> more detailed scribe documentation is at https://
niklasl: how to cope with questions on named graphs
tl: we've had discussions on named graphs; it's good to give an overview on status
AndyS: niklasl still has named graph example in current slides
<TallTed> s/s\/niklasl: only\///
niklasl: will leave that example in there, just to illustrate
niklasl: keen to keep it; but, may change my mind
niklasl: I can try and rehearse the talk now
ora: we can individually read the slides now and then discuss
<tl> https://
everybody is reading https://
<AndyS> Part 2: https://
<gb> CLOSED Issue 7 test *again* ghurlbot configuration (by ghurlbot)
<Zakim> pchampin, you wanted to comment on "reifiying predicate" S3 and to comment on "subgraph" in S3 of part 2
pchampin: on slide 3; you note a reifying predicate; consider putting in quotes as the naming is still under discussion
pchampin: in second presentation, issue re reification in named graphs on slide 3
link for second presentation: https://
pchampin: propose to word it differently
pchampin: would refrain from calling it a subgraph
AndyS: makes another suggestion to reword it
niklasl: will consider the suggestion
<Zakim> ora, you wanted to comment on ":source" on S14
ora: part 1, slide 14: code has more stuff than the diagram (source)
<Zakim> tl, you wanted to comment on slide #7 in part 1
<gb> CLOSED Issue 7 test *again* ghurlbot configuration (by ghurlbot)
tl: has fundamental issue; this is only a reifier of a possible truth ...
tl: ... diagram thus needs two parts
niklasl: diagram is meant to reflect both
tl: still seems incorrect; would suit the "old" rdf*
ora: how about putting assertion first? also, replace "truth" with "assertion"
TallTed: how about just "triple"
niklasl: that would be ambiguous
ora: part of problem is that "semantics" of diagrams are unclear
<pchampin> well, the abstract syntax and the concrete syntax (Turtle) are pretty much figured out, so we can rely on the code in the slides.
pchampin: it's the same abstract syntax, though.
ora: idea for diagrams: if we have a reifier, always have oval around reified triple
ora: use color to indicate asserted or not
<pchampin> or dashed vs. solid arrows
<pchampin> ora, you proposal breaks as soon as we have different reifiers
ora: clarify that we don't refer to to constituent parts
<TallTed> I think `<< :Alice :bought :LennyTheLion >> .` is problematic. It's not a statement. it's just a bare "noun". `<< :Alice :bought :LennyTheLion >>` (no terminal `.`) is less troublesome.
<doerthe> +1 to TallTed
ora: if we add too much nuance, then we will lose a lot of the audience
<TallTed> (and `statement` might be better than `truth`/`assertion` in title of slide 7)
<pchampin> TallTed, doerthe, it *is* a statement, though: it says "there is a reifier for (this triple)".
AndyS: comments on consistency of diagrams
<doerthe> ok pchampin, what is it syntactic sugar for then?
AndyS: unclear what reifier points to (a part of the triple)?
<pchampin> doerthe this is equivalent to what you have in slide 5. I agree this is odd, but it does make sense
<pchampin> think of it as [ rdf:reifies <<( S P O )>> ] .
<TallTed> pchampin -- then *that* is what should be written.
niklasl: could make things even more complicated
<doerthe> ah, it is the other way around, I thought the bare triple term would be the <<...>> one, but it is the <<(...)>>
<pchampin> you can always replace << S P O >> with [ rdf:reifies <<( S P O )>> ]. Either on its own OR as the subject/object of another triple.
<doerthe> maybe something to say on the slide for others like me?
ora: perhaps accept that diagrams are more illustrative
<doerthe> pchampin, I understand, but mixed up the two representations
ora: can gloss over distinction between triple & triple term
AndyS: better transition between slides 6 & 7
I will have to leave in around 15-20 minutes as well
<doerthe> ok, that was bad, if PA leaves, we all go
<tl> will somebody restart it?
<TallTed> just rejoin zoom... it's because pchampin, the host, left the room.
tl: not satisfied with the slides for now
tl: same point as ora, i.e., have reifier point to ellipse instead of predicate
tl: this is an essential part of our solution - reifier + triple
tl: central problem = cannot refer to triple as it is asserted
tl: suggestion on transition from slide 6 to 7
ora: apart from diagram, other comments?
Souri: reifier should point to the whole oval on slides 5, 6
Souri: triples are unrelated aside from the fact they share the same SPO
william_vw: perhaps drop bullet 2 from slide 3
niklasl: will delete it from slide
niklasl: triple term is reference to propositional atom on slides
tl: should still explicate that these are 2 separate things
tl: points to fundamental disagreement on rdf:states
niklasl: will attempt to update slide 7 + verbal wording
<Souri> My understanding: The [] rdf:reifies <<( :s :p :o )>> triple should diagrammatically be represented as a blanknode pointing to a shaded s-p-o oval. A separate s-p-o should appear in the diagram for the asserted triple :s :p :o .
<doerthe> maybe do 6 and add the triple to the diagram (maybe with dotted lines)
<tl> Souri +1
continuing discussion on diagram representation
ora: oval around the triple is the "triple term"
ora: e.g., dotted lines to indicate asserted nature (as doerthe said)
ora: then, no repetition needed
tl: need more examples in slides
<Souri> The diagram in Slide 7 should ideally be: the diagram in Slide 6 + a new open s-p-o (not inside an oval). Then, we can, as a shortcut/convenience, show a compact diagram which has a non-shaded oval (or a dotted oval) to indicate it represents both the triples (shown in N-triple).
tl: not glad with one of the examples
niklasl: don't want to put animations in slides
niklasl: at least, will put oval around triple on slide 7
niklasl: still want to keep visuals close to syntax
AndyS: who is the Tuesday presentation aimed at?
AndyS: is it open to the general public?
Souri: triple & triple term simply happen to share the same SPO on slide 7
<tl> Proposal for Slide 7: "Reifier of a _possible_ Truth"
Souri: so, really 2 different things
Souri: usually we combine them, but still different things
tl: simply move graphic from slide 6 to slide 7 to make separation clear